Mobley v. Workday, Inc.

[PROPOSED] ORDER ADOPTING WORKDAY’S
STRIKES TO THE TAC [ 3:23-CV-00770-RFL]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
JULIE A. TOTTEN (STATE BAR NO. 166470)
jatotten@orrick.com
ERIN M. CONNELL (STATE BAR NO. 223355)
econnell@orrick.com
KAYLA D. GRUNDY (STATE BAR NO. 300513)
kgrundy@orrick.com
ALEXANDRIA R. ELLIOTT (STATE BAR NO. 320293)
aelliott@orrick.com
ORRICK, HERRINGTON & SUTCLIFFE LLP
The Orrick Building
405 Howard Street
San Francisco, CA 94105-2669
Telephone: (415) 773-5700
Facsimile: (415) 773-5759
Attorneys for Defendant
WORKDAY, INC.
UNITED STATES DISTRICT COURT
NORTHERN DISTRICT OF CALIFORNIA
DEREK L. MOBLEY, FAITHLINH ROWE;
JILL E. HUGHES, AND SHEILAH
JOHNSON-ROCHA, on behalf of themselves
and all others similarly situated,
Plaintiff,
v.
WORKDAY, INC.,
Defendant.
Case No. 3:23-cv-00770-RFL
[PROPOSED] ORDER GRANTING
WORKDAY’S SUBMISSION OF
PARTIALLY STRICKEN THIRD
AMENDED COMPLAINT AS
MODIFIED
Courtroom: 4 (15th Floor)
Judge: Hon. Rita F. Lin
Third Amended Complaint Filed: March 27,
2026
Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 1 of 59

1 [PROPOSED] ORDER ADOPTING WORKDAY’S
STRIKES TO THE TAC [ 3:23-CV-00770-RFL]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
[PROPOSED] ORDER AS MODIFIED
On June 22, 2026, the Court issued an Order granting in part and denying in part Defendant
Workday, Inc.’s Motion to Dismiss and to Strike Plaintiffs’ Third Amended Complaint. See ECF.
No. 360. The Court directed Workday to file a proposed order narrowing its proposed strikes to the
Third Amended Complaint consistent with that Order. By filing its proposed strikes consistent with
the Court’s previous Order, Workday does not concede any of Plaintiffs’ allegations are true or
proper and the filing of such a proposal shall not be used against Workday for any purpose in this
litigation. Having reviewed Workday’s submission, the Court finds that the proposed strikes, as
modified, are consistent with the June 22, 2026 Order and hereby ADOPTS AS MODIFIED
Workday’s proposed strikes to the Third Amended Complaint. The allegations that Rowe applied
to Workday through “the same automated platform” will not be struck, because they support
Rowe’s claim that she was discriminated against by Workday’s AI tools. An individual would be
a member of the putative class even if the only employer they applied to was Workday, so long as
they were “subjected to the challenged discriminatory screening process.”
Plaintiffs shall file a clean Fourth Amended Complaint consistent with the stricken
complaint filed herewith. Defendants’ answer to the Fourth Amended Complaint shall be due by
July 15, 2026.
IT IS SO ORDERED.
Dated: July 1, 2026 By:
Hon. Rita F. Lin
UNITED STATES DISTRICT COURT
JUDGE
Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 2 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
PLAINTIFFS’ THIRD AMENDED COMPLAINT
Lee D. Winston 	Robert L. Wiggins, Jr.
Admitted Pro Hac Vice 	Admitted Pro Hac Vice
lwinston@winstoncooks.com 	rwiggins@wigginschilds.com
Roderick T. Cooks 	Ann K. Wiggins
Admitted Pro Hac Vice 	Admitted Pro Hac Vice
rcooks@winstoncooks.com 	awiggins@wigginschilds.com
Bethany Mae Logan 	Samuel Fisher, Jr.
Admitted Pro Hac Vice 	Admitted Pro Hac Vice
bneal@winstoncooks.com 	sf@wigginschilds.com
Winston Cooks, LLC 	Jennifer Wiggins-Smith
420 20th Street North, Suite 2200 	Admitted Pro Hac Vice
Birmingham, AL 35203 	jsmith@wigginschilds.com
Nicki Leili Lawsen
Local Counsel: 	Admitted Pro Hac Vice
Jay Greene, JD, CPA 	nlawsen@wigginschilds.com
Greene Estate, Probate, and Elder 	Freddrick Malik Moore
Law Firm 	Admitted Pro Hac Vice
447 Sutter Street, Suite 435 	mmoore@wigginschilds.com
San Francisco, CA 94108
Attorneys for the Plaintiff and Proposed Class and Collective Members
UNITED STATES DISTRICT COURT
FOR THE NORTHERN DISTRICT OF CALIFORNIA
DEREK L. MOBLEY, JILL E. HUGHES,
SHEILAH JOHNSON-ROCHA, and
FAITHLINH ROWE. on behalf of themselves
and all others similarly situated,
Plaintiffs,
v.
WORKDAY, INC.,
Defendant.
Case No. 3:23-cv-00770-RFL
PLAINTIFFS’ THIRD
AMENDED COMPLAINT
NATURE OF COMPLAINT
Plaintiffs Derek L. Mobley (“Mobley”), Jill E. Hughes (“Hughes”), Sheilah Johnson-Rocha
(“Johnson-Rocha”), and FaithLinh Rowe (“Rowe”) (collectively, “Representative Plaintiffs”) bring
this action on behalf of themselves and all others similarly situated for injunctive, declaratory, and
monetary relief against Defendant Workday, Inc. (“Workday” or “Defendant”). Plaintiffs challenge
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 1 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 1 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 3 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
2
PLAINTIFFS’ THIRD AMENDED COMPLAINT
Defendant’s pattern or practice of discrimination on the basis of race—including African American
and Asian American applicants—sex (female), age (over 40), and disability, in violation of Title
VII of the Civil Rights Act of 1964, the Age Discrimination in Employment Act of 1967
(“ADEA”), the Americans with Disabilities Act Amendments Act of 2008 (“ADAAA”), and the
California Fair Employment and Housing Act (“FEHA”), Cal. Gov. Code § 12940 et seq.
This case arises from Defendant’s use of artificial intelligence (“AI”), machine learning
(“ML”), and other algorithmic decision-making tools to screen, evaluate, rank, and reject job
applicants, including, but not limited to, applicants for jobs and employment opportunities in
California with Workday and other employers. These tools are not passive or ministerial. They
function as decision-making systems that materially influence, and in many cases determine,
whether an applicant is advanced or rejected for employment opportunities.
Workday provides human resource management and applicant-screening services to
medium-sized and large organizations across numerous industries, including professional services,
financial services, healthcare, education, government, technology, media, retail, and hospitality.
Employers subscribe to Workday’s platform to manage hiring workflows, including the intake,
processing, scoring, ranking, and disposition of job applications. 	Workday utilizes those same
practices and application platforms for applications seeking employment as Workday employees.
At the scale at which Workday operates, its platform serves as a primary gateway to employment.
Workday has processed millions of job requisitions and hundreds of millions of applications
annually. In May 2023 alone, Workday processed approximately 2.2 million job requisition
transactions—representing nearly 22% of all job openings in the United States during that period.
As a result, Workday’s system does not merely assist hiring decisions; it structures access
to employment opportunities across the national labor market.
Defendant’s discriminatory conduct arises from its policy and practice of using automated,
algorithm-driven systems—in lieu of individualized human review—to evaluate and screen
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 2 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 2 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 4 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
3
PLAINTIFFS’ THIRD AMENDED COMPLAINT
applicants at scale. These systems are trained on historical data and prior hiring outcomes and are
designed to identify patterns and relationships between applicant characteristics and hiring
decisions. Once deployed, the system applies those learned relationships to new applicants in order
to predict whether they should be advanced or rejected. Workday uses those same automated,
algorithm-driven systems for its own applicants for jobs as Workday employees. In California and
elsewhere.
Because these systems rely on historical data and statistical modeling, they reproduce and
amplify disparities embedded in prior hiring decisions and labor market structures. Bias may enter
the system through multiple pathways, including the composition of training data, the selection and
weighting of input variables, the design of model objectives, and the evaluation criteria used to
assess candidate “fit.”
Even where protected characteristics such as race, age, sex, or disability are not explicitly
provided to the system, they can be inferred through correlated variables. For example, employment
gaps may reflect disability or health conditions, years of experience may serve as a proxy for age,
educational background and institutional affiliations may correlate with race, and other application
data may reflect gender or caregiving responsibilities. Through these proxy variables, the system is
capable of reproducing discriminatory outcomes without explicitly classifying applicants by
protected characteristics.
Once deployed, Defendant’s system operates uniformly across all applicants. Each applicant
is subjected to the same underlying model, scoring mechanisms, and decision-making logic. As a
result, any bias embedded in the system manifests systematically, producing consistent disparities in
outcomes across protected groups rather than isolated or individualized instances of discrimination.
The discriminatory effects of such systems are measurable at the population level, including
through disparities in selection rates, rejection rates, ranking outcomes, and other screening decisions
across demographic groups. Because the same system is applied across employers and job postings,
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 3 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 3 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 5 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
4
PLAINTIFFS’ THIRD AMENDED COMPLAINT
these disparities are attributable to Defendant’s centralized platform rather than individualized
employer discretion.
The experiences of the Representative Plaintiffs illustrate the operation and impact of
Defendant’s system. Plaintiff Mobley, an African American male over the age of forty with
disabilities, applied for numerous positions through Workday’s platform and was repeatedly rejected
despite being qualified. Plaintiff Hughes, a woman over forty with physical disabilities including
cancer and asthma, similarly experienced repeated rejections, including rejections that occurred
rapidly and without individualized review. Plaintiff Johnson-Rocha, an African American woman
over forty, submitted approximately 2,000 applications through systems utilizing Workday and was
consistently rejected, including for positions that were reposted and for which she remained qualified.
Plaintiff Rowe, an Asian American woman over the age of forty and a qualified professional, likewise
experienced repeated and often immediate rejections through Workday’s system, including for
positions associated with Defendant itself.
The uniformity, speed, and consistency of these rejections across different employers, job
postings, and time periods are indicative of automated decision-making rather than individualized
human review. These experiences are consistent with a centralized system applying standardized
criteria and algorithmic decision-making processes across all applicants.
Defendant’s system disproportionately excludes applicants who are African American, Asian
American, female, over the age of forty, and/or disabled. These disparities are not the result of chance
or isolated decision-making, but instead arise from the structure, design, and operation of
Defendant’s algorithmic screening tools.
As a direct and proximate result of Defendant’s conduct, Plaintiffs and the proposed classes
have been denied employment opportunities, lost wages and career advancement, and suffered
emotional distress. Plaintiffs and class members have also been deterred from applying to positions
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 4 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 4 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 6 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
5
PLAINTIFFS’ THIRD AMENDED COMPLAINT
that utilize Workday’s platform due to the futility of repeated rejections generated by Defendant’s
system.
Defendant’s conduct is ongoing, systemic, and unlawful. Plaintiffs seek injunctive relief to
halt Defendant’s discriminatory practices, declaratory relief to establish their unlawfulness, and
monetary damages to compensate Plaintiffs and the proposed classes for the harms they have
suffered.
JURISDICTION AND VENUE
1. 	The jurisdiction of this Court is invoked pursuant to 28 U.S.C. §§ 1331, 1343(3), and
(4),2201 and 2202, 42 U.S.C. 2000d-2 and 2000e5(f), and 29 U.S.C. § 621, et seq. Supplemental
jurisdiction for the state law claims are invoked pursuant to 28 U.S.C. §1367.
2. 	This is a suit authorized and instituted pursuant to the Act of Congress known as
Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000 et seq., as amended, the Age
Discrimination in Employment Act of 1967, 29 U.S.C. § 621, et seq., and the ADA Amendments
Act of 2008 (“ADAAA”).
3. 	Venue is proper in the Northern District of California pursuant to 28 U.S.C. § 1391(b)
and (c) because Defendant Workday, Inc. maintains its principal place of business in Pleasanton,
California, within this District, and a substantial part of the events or omissions giving rise to
Plaintiffs’ claims occurred in this District.
4. 	Defendant’s applicant-screening systems—including its artificial intelligence and
machine-learning tools used to evaluate, rank, and reject applicants—are designed, developed,
maintained, and controlled from its California headquarters.
PARTIES
5. 	Plaintiff Derek L. Mobley is an African American male over the age of forty (40) who
suffers from depression and anxiety, conditions that constitute disabilities within the meaning of the
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 5 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 5 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 7 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
6
PLAINTIFFS’ THIRD AMENDED COMPLAINT
ADAAA. Mobley is a job applicant who has applied for numerous positions through Defendant’s
platform and has been subjected to Defendant’s automated screening system.
6. 	Plaintiff Jill E. Hughes is a female over the age of forty (40) and an individual with
disabilities within the meaning of the ADAAA, including physical impairments such as cancer and
asthma. Hughes is a qualified professional who has applied for positions through Defendant’s
platform and has been subjected to Defendant’s automated screening tools, which disproportionately
exclude female, disabled, and over-forty applicants. Beginning on or about April 15, 2024, Hughes
experienced repeated rejections consistent with the operation of Defendant’s algorithmic screening
system. Hughes has opted into the age-based collective action. [See Dkt. 103-1].
7. 	Plaintiff Sheilah Johnson-Rocha is an African American female over the age of forty
(40). Johnson-Rocha is a qualified professional who has applied for positions through Defendant’s
platform and has been subjected to Defendant’s automated screening tools, which disproportionately
exclude African American, female, and over-forty applicants. On or about September 3, 2024,
Johnson-Rocha experienced rejections consistent with the operation of Defendant’s system. Johnson-
Rocha has opted into the age-based collective action. [See Dkt. 103-1].
8. 	Plaintiff FaithLinh Rowe is an Asian American female over the age of forty (40) and
a qualified HRIS professional with extensive human resources experience. Rowe has applied for
multiple positions through Defendant’s platform, including positions located in California with
Defendant Workday, Inc. itself, and has been subjected to Defendant’s automated screening tools,
which disproportionately exclude female and over-forty applicants and, upon information and belief,
adversely impact applicants across multiple racial groups, including Asian American applicants.
Rowe has opted into the age-based collective action. [See Dkt. 122-1].
9. 	In addition to using the challenged recruitment and hiring procedures in screening and
selecting its own employees, Defendant Workday, Inc. acts as an agent of its client-employers by
exercising delegated authority over the hiring process, including screening, ranking, and filtering
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 6 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 6 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 8 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
7
PLAINTIFFS’ THIRD AMENDED COMPLAINT
applicants through its platform. Employers rely on the results of Workday’s selection procedures in
making hiring decisions. In the alternative, Workday functions as an indirect employer because it
controls access to employment opportunities by determining which applicants are advanced or
rejected through its system.
CONDITIONS PRECEDENT TO SUIT UNDER
TITLE VII, THE ADEA AND THE ADAAA
10. 	On June 3, 2021, Plaintiff Mobley filed a charge of discrimination with the Oakland
Field Office of the United States Equal Employment Opportunity Commission (“EEOC”), and on
July 19, 2021, he filed an amended charge. On November 22, 2022, the EEOC issued Mobley a
Dismissal and Notice of Right to Sue. Mobley timely filed this action within ninety (90) days of
receipt of that notice.
11. 	On May 3, 2024, Plaintiff Jill E. Hughes filed a charge of discrimination with the
United States Equal Employment Opportunity Commission (“EEOC”), and on May 13, 2025, the
EEOC issued Hughes a Dismissal and Notice of Right to Sue. Hughes timely joined this action within
ninety (90) days of receipt of that notice.
12. 	Hughes has therefore satisfied all administrative prerequisites to suit under Title VII,
the ADEA, and the ADAAA.
13. 	On 	September 	3, 	2024, 	Plaintiff 	Sheilah 	Johnson-Rocha 	filed 	a 	charge 	of
discrimination with the United States Equal Employment Opportunity Commission (“EEOC”), and
on May 13, 2025, the EEOC issued Johnson-Rocha a Dismissal and Notice of Right to Sue. Johnson-
Rocha timely joined this action within ninety (90) days of receipt of that notice.
14. 	Johnson-Rocha has therefore satisfied all administrative prerequisites to suit under
Title VII and the ADEA.
15. 	On May 9, 2025, Plaintiff FaithLinh Rowe filed a charge of discrimination with the
United States Equal Employment Opportunity Commission (“EEOC”), which was dual-filed with
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 7 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 7 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 9 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
8
PLAINTIFFS’ THIRD AMENDED COMPLAINT
the California Civil Rights Department (“CRD”), alleging discrimination on the basis of race, sex,
and age in violation of Title VII of the Civil Rights Act of 1964 and the Age Discrimination in
Employment Act of 1967. On or about August 13, 2025, the EEOC issued Rowe a Dismissal and
Notice of Rights.
16. 	Rowe has satisfied all administrative prerequisites to suit under federal law and has
exhausted her administrative remedies under the California Fair Employment and Housing Act
(“FEHA”).
CLASS ACTION ALLEGATIONS
17. 	The Representative Plaintiffs bring this action individually and on behalf of all others
similarly situated under Rule 23 of the Federal Rules of Civil Procedure and seek certification of the
following subclasses:
•All African-American applicants or former applicants who from September 24, 2020,
to the present were subjected to the challenged discriminatory screening process.
*All Asian-American applicants or former applicants who from September 24, 2020,
to the present were subjected to the challenged discriminatory screening process.
•All female applicants or former applicants who from November 8, 2023, to the
present were subjected to the challenged discriminatory screening process.
•All applicants or former applicants over the age of forty (40) who from September
24, 2020, to the present were subjected to the challenged discriminatory screening
process.
•All applicants or former applicants who, from September 24, 2020, to the present,
have a disability or medical condition affecting mental, cognitive, or physical
functioning, 	and 	who 	were 	evaluated, 	screened, 	ranked, 	or 	rejected 	through
Defendant’s automated assessment tools, including cognitive tests, personality tests,
or other algorithmic screening mechanisms.
Representative Plaintiffs in the case at bar challenge systemic discrimination by, and seek
class-wide relief against, Workday for its utilization of discriminatory screening tools as part of its
employment policies and procedures which constitute a pattern and practice of discrimination on the
basis of race, sex, age, and disability with respect to selections. These screening tools have been
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 8 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 8 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 10 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
9
PLAINTIFFS’ THIRD AMENDED COMPLAINT
continuously utilized by the Defendant since at least 2017, and their implementation and use has
personally harmed the named Representative Plaintiffs, and the putative class members they seek to
represent. Workday’s client-customers delegate to it the hiring process, recruitment, and onboarding
of employees. Workday then utilizes screening tools, to include Workday branded assessments
and/or tests, to process and interpret an applicant’s qualifications and recommend whether the
applicant should be accepted or rejected.
Workday’s utilization of these screening tools relies upon subjective practices which have
caused disparate impact and disparate treatment to applicants who are African-American, Asian-
American, women, over the age of forty (40) or and/or disabled. Applicants who are not members of
these protected groups and who are similarly situated to the Representative Plaintiffs, have not been
subjected to such discriminatory treatment.
A. 	Common Questions of Law and Fact
18. 	The prosecution of the claims of the Representative Plaintiffs requires adjudication of
numerous questions of law and fact common to his individual claims and those of the putative classes
he seeks to represent. The common questions of law would include, inter alia: (a) whether the
Defendant’s screening products cause African-American, Asian-Americans, women, individuals
over the age of forty (40), and/or individuals with a disability to be disproportionately and more
likely denied employment; (b) whether the Defendant’s doing so cannot be justified as a necessary
business practice for evaluating potential employees; and (c) whether the Defendant’s screening
products have a disparate impact on applicants who are African-American, Asian-American, women,
over the age of forty (40), and/or disabled in violation of Title VII of the “Civil Rights Act of 1964,”
42 U.S.C. § 2000 et seq., the Age Discrimination in Employment Act of 1967, 29 U.S.C. § 621, et
seq., and the ADA Amendments Act of 2008 (“ADAAA”).
17. 	The common questions of fact would include, inter alia: (1) whether Workday’s
administration of its screening products discriminated against the aforementioned applicants because
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 9 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 9 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 11 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
10
PLAINTIFFS’ THIRD AMENDED COMPLAINT
of their race, sex, age, and/or disability with regards to hiring; (2) whether compensatory and punitive
damages, injunctive relief, and other equitable remedies for the class are warranted; and (3) whether
Workday discriminated against the aforementioned protected groups in other terms and conditions
of employment. The details of the Representative Plaintiffs’ claims are encompassed within the
claims prosecuted on behalf of the class and set forth in this Complaint.
B. 	Typicality
19. 	The claims of the Representative Plaintiffs are typical of those of the members of
the class. The Representative Plaintiffs and all class members have been and are similarly adversely
affected by the systemic discriminatory practices complained of herein. Specifically, the
representative claims, like those of the class members, arise out of Defendant's pervasive
discriminatory conduct in hiring and other terms and
conditions of employment.
20. 	The relief necessary to remedy the claims of the Representative Plaintiffs is the same
relief that is necessary to remedy the claims of the putative class members in this case. The
Representative Plaintiffs seek the following relief for individual claims and class claims asserted
herein: (1) declaratory judgment that Defendant has engaged in systemic discrimination against
African-Americans, Asian-Americans, women, individuals over the age of forty (40), and/or the
disabled; (2) a permanent injunction against such continuing discrimination; (3) injunctive relief
which reforms Workday’s screening products, policies, practices and procedures so that the
Representative Plaintiff and the class members will be able to compete fairly in the future for jobs
and enjoy terms and conditions of employment traditionally afforded similarly situated employees
outside of the protected categories; (4) backpay, front pay, compensatory damages, and other
equitable remedies necessary to make the Plaintiff, and the class, whole from Workday’s past
discrimination; and, (5) attorneys' fees, costs, and expenses.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 10 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 10 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 12 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
11
PLAINTIFFS’ THIRD AMENDED COMPLAINT
C. 	Numerosity and Impractability Of Joinder
21. 	The class that the Representative Plaintiffs seek to represent is too numerous to make
joinder practicable. The proposed class consists of numerous former, current, and future applicants
who have been denied employment due to the discriminatory administration of Workday’s screening
products. Workday’s pattern or practice of discrimination also makes joinder impracticable by
making it impractical and inefficient to identify many members of the class prior to the determination
of the merits of Workday’s class wide liability. Thus, the number of class members is currently
indeterminate but is certainly numerous.
D. 	Adequacy of Representation
22. 	The Representative Plaintiffs will fairly and adequately protect the interests of the
class inasmuch as they are broadly representative, as reflected in the preceding paragraphs. There are
no conflicts of interest present with the members of the proposed class as each would benefit from
the imposition of a remedy for the Defendant’s discriminatory employment practices. The
Representative Plaintiffs have retained counsel experienced in litigating major class actions in the
field of employment discrimination, and who are prepared and able to meet the time and fiscal
demands of class action litigation of this size and complexity. The combined interest, experience,
and resources of the Representative Plaintiffs and their counsel to litigate competently the individual
and class claims of employment discrimination at issue satisfy the adequacy of representation
requirement under Fed.R.Civ.P. 23(a)(4).
E. 	Efficiency of Class Prosecution of Common Claims
23. 	Certification of a class of similarly-situated applicants is the most efficient and
economical means of resolving the questions of law and fact that are common to the individual claims
of the Representative Plaintiffs and the proposed class. The individual claim of the Representative
Plaintiffs require resolution of the common question of whether Defendant has engaged in a systemic
pattern of discrimination against African-Americans, Asian-Americans, women, those over forty
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 11 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 11 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 13 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
12
PLAINTIFFS’ THIRD AMENDED COMPLAINT
(40) and the disabled. The Representative Plaintiffs seek remedies to undo the adverse effects of such
discrimination in his own life and career. The Representative Plaintiffs have standing to seek such
relief because of the adverse effect that such discrimination has had on him individually and on the
putative classes he seeks to represent, in general. To gain such relief for themselves, as well as for
the putative class members, the Representative Plaintiffs will first establish the existence of systemic
discrimination as the premise of the relief they seek. Without class certification, the same evidence
and issues would be subject to re-litigation in a multitude of individual lawsuits with an attendant
risk of inconsistent adjudications and conflicting obligations. Certification of the subclasses affected
by the common questions of law and fact is the most efficient and judicious means of presenting the
evidence and arguments necessary to resolve such questions for the Representative Plaintiffs, the
classes and the Defendant. The Representative Plaintiffs’ individual and class claims are premised
upon the traditional bifurcated method of proof and trial for systemic disparate treatment/impact
claims of the type at issue in this complaint. Such a bifurcated method of proof and trial is the most
efficient method of resolving such common issues.
F. 	Certification is Sought Pursuant to Fed.R.Civ. P. 23(b)(2)
24. 	Workday has acted on grounds generally applicable to the Representative Plaintiffs
and the proposed class by adopting and following systemic practices and procedures that discriminate
on the basis of race, sex, age, and/or disability. Workday’s screening products are regularly used to
discriminate on the basis of race, sex, age, and/or disability. Workday has refused to act on grounds
generally applicable to the putative class by: (1) refusing to adopt or follow screening productions
and selection procedures which do not systemically discriminate on the basis of race, sex, age, and/or
disability. Workday’s discriminatory screening products have made appropriate final injunctive relief
and declaratory relief with respect to the class as a whole. The injunctive relief and declaratory relief
are the predominate reliefs sought because they are both the cumulation of the proof of the
Defendant’s individual and class-wide liability at the end of Stage I of a bifurcated trial and the
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 12 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 12 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 14 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
13
PLAINTIFFS’ THIRD AMENDED COMPLAINT
essential predicate for the Representative Plaintiffs and the classes members entitlement to monetary
and non-monetary remedies at Stage II of such a trial. Declaratory and injunctive relief flow directly
and automatically from proof of the common questions of law and fact regarding the existence of
systemic discrimination against individuals on the basis of race, sex, age, and/or disability. Such
relief is the factual and legal predicate for the Representative Plaintiffs and the class members
entitlement to injunctive and equitable remedies caused by such systemic discrimination
G. 	Alternatively, Certification is Sought Pursuant to Fed. R.Civ. P. 23(b)(3)
25. 	The common issues of fact and law affecting the claims of the Representative
Plaintiffs and proposed class members, including, but not limited to, the common issues identified
above, predominate over any issues affecting only individual claims. A class action is superior to
other available means for the fair and efficient adjudication of the claims of the Representative
Plaintiffs and members of the proposed classes. The cost of proving the Defendant’s pattern or
practice of discrimination makes it impracticable for the named Plaintiffs and members of the
proposed class to control the prosecution of their claims individually. The Northern District of
California is the most logical forum in which to litigate the claims of the Representative Plaintiffs
and the proposed class in this case because the Defendant’s home office is here and it engages in or
ratifies illegal conduct adversely affecting the Plaintiff here.
H. 	Alternatively, Certification is Sought Pursuant to Fed. R. Civ. P. 23(c)(4)
for Injunctive and Declaratory Relief
26. 	Alternatively, claims for injunctive and declaratory relief for the Injunctive Relief
Class are properly certified under Federal Rule of Civil Procedure 23(c)(4) because such claims
present only common issues, the resolution of which would advance the interests of the parties in an
efficient manner.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 13 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 13 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 15 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
14
PLAINTIFFS’ THIRD AMENDED COMPLAINT
I. 	Alternatively, Certification is Sought Pursuant to Fed. R. Civ. P. 23(c)(4)
For Class Wide Liability
27. 	Alternatively, class wide liability claims are properly certified under Federal Rule of
Civil Procedure 23(c)(4) for the Classes because such claims present only common issues, the
resolution of which would advance the interests of the parties in an efficient manner.
J. 	Punitive Damages May Alternatively be certified Pursuant to Fed. R. Civ.
23(b)(2)
28. 	Punitive damages liability may alternatively be certified under Federal Rule of Civil
Procedure 23(b)(2) because such relief focuses on the conduct of Workday and not the individual
characteristics of the Plaintiff and are an allowable form of incidental monetary relief.
CLAIMS OF REPRESENTATIVE PLAINTIFFS
A. 	Derek Mobley
29. 	Since at least 2017, Plaintiff Mobley has applied for over one hundred (100) positions
that utilized Defendant Workday, Inc.’s platform as the gateway for applicant screening and hiring.
Each application resulted in rejection.
30. 	Workday’s 	platform 	is 	not 	a 	marginal 	tool—it 	is 	a 	dominant 	labor-market
intermediary. Workday services more than 10,000 organizations, including over half of Fortune 500
companies, and processes a substantial portion of U.S. job applications. As a result, Workday
functions as a centralized screening infrastructure governing access to employment opportunities.
31. 	Mobley’s application process followed a uniform, Workday-controlled pathway. He
would respond to job postings on third-party platforms such as LinkedIn, Indeed, or Monster, which
then redirected him to employer-specific Workday portals.
32. 	Upon information and belief, these portals—e.g., “@myworkday.com” domains—are
standardized interfaces through which Workday collects, structures, and evaluates applicant data
using common backend systems.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 14 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 14 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 16 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
15
PLAINTIFFS’ THIRD AMENDED COMPLAINT
33. 	To proceed, Mobley was required to create login credentials, upload his résumé, and
submit structured data inputs dictated by Workday’s system architecture.
34. 	Mobley’s résumé reflects graduation from Morehouse College, a historically Black
college, and extensive work history—information that can function as proxies for protected
characteristics such as race and age when processed through data-driven models.
35. 	Numerous applications required Mobley to complete Workday-branded assessments
and/or personality tests as a condition of consideration.
36. 	Upon information and belief, these assessments constitute standardized components
of Workday’s screening system and are used across multiple employers without material variation.
37. 	These assessments are not merely neutral questionnaires. Rather, they are data-
collection instruments designed to generate behavioral and psychological profiles that are fed into
algorithmic decision-making tools.
38. 	Upon information and belief, such tools operate as “classification systems” that sort
applicants based on predicted suitability, using correlations derived from historical data rather than
job-related causation.
39. 	Mobley suffers from depression and anxiety. The assessments required by Workday
are likely to reveal or correlate with such conditions and thus function as disability-related inquiries
or medical examinations under the ADA.
40. 	Individuals with such conditions are more likely to perform worse on these
assessments, resulting in systematic exclusion at the screening stage.
41. 	Mobley’s experience reflects a consistent and highly probative pattern of near-
immediate rejection following submission of applications through Workday’s platform, strongly
indicating automated screening rather than individualized review.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 15 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 15 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 17 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
16
PLAINTIFFS’ THIRD AMENDED COMPLAINT
42. 	For example, on or about September 23, 2018, Mobley applied for a Corporate Travel
Consultant position with Expedia via the Workday platform. On October 2, 2018, at approximately
2:19 a.m., he received a rejection email.
43. 	On March 31, 2019, Mobley applied for a Claim Support Representative position with
Fiserv via Workday. The following morning, April 1, 2019, at approximately 9:40 a.m., he received
a rejection email—less than one day after applying.
44. 	In June 2019, Mobley applied for a Help Desk Support Technician position with NCR
Corporation via Workday and was rejected on June 20, 2019, despite meeting the stated
qualifications.
45. 	On August 31, 2019, Mobley applied for an Associate Customer Care Specialist
position with Duke Energy via Workday, completed a required Workday-branded assessment, and
was subsequently rejected without explanation.
46. 	On that same date, August 31, 2019, Mobley applied for a Customer Service
Representative position with Unum via Workday and was rejected that same day at approximately
12:52 a.m., reflecting an almost instantaneous automated disposition.
47. 	On January 29, 2022, at approximately 12:55 a.m., Mobley applied for a Customer
Services Specialist position with Unum via Workday. Less than one hour later, at approximately 1:50
a.m., he received a rejection email.
48. 	On January 9, 2024, Mobley applied for a Customer Support Representative position
with ResMed via Workday. On January 11, 2024, at approximately 3:52 a.m., he received a rejection
email.
49. 	Across these and numerous other applications, Mobley was rejected despite meeting
or exceeding the stated qualifications for the positions.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 16 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 16 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 18 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
17
PLAINTIFFS’ THIRD AMENDED COMPLAINT
50. 	The timing of these rejections—often occurring within hours or even minutes of
submission, including during overnight hours—supports a reasonable inference that Workday’s
algorithmic decision-making tools automatically screened out Mobley’s applications.
51. 	This 	pattern 	is 	consistent 	across 	multiple 	employers, 	industries, 	and 	years,
demonstrating that the decision-making mechanism is centralized within Workday’s platform rather
than individualized by each employer.
52. 	Upon information and belief, these tools utilize machine learning models whose
behavior depends on trained parameters, scoring thresholds, and data interactions, such that their
discriminatory effects manifest in outcomes rather than observable rules.
B. 	Jill E. Hughes
53. 	Ms. Jill E. Hughes is a white female over the age of forty (40), born on February 4,
1974. Ms. Hughes is a cancer survivor and currently suffers from asthma—serious medical
conditions that qualify as disabilities under the ADA and that may affect respiratory function,
stamina, and work patterns.
54. 	Ms. Hughes holds a B.A. from the University of Cincinnati and an M.A. from the
University of Chicago and possesses more than twenty-eight (28) years of executive-level experience
as a Senior Medical Writer, Project Manager, and Scientific Director. Her experience includes high-
level work in the corporate, medical, and media sectors, and she is well-qualified for the positions to
which she applied.
55. 	Since at least December 2023, Ms. Hughes has applied for hundreds of positions
through hiring systems powered by Defendant Workday, Inc., which function as the primary gateway
for sourcing, recruiting, and applicant screening. 	These applications include positions with major
employers such as 3M, Cigna, Veralto, Stryker, Becton Dickinson, FIS Global, General Electric,
Huntington, and Parsons Corporation. Despite meeting or exceeding the stated qualifications for
these roles, Ms. Hughes has been rejected in every instance.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 17 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 17 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 19 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
18
PLAINTIFFS’ THIRD AMENDED COMPLAINT
56. 	Upon information and belief, Workday’s algorithmic decision-making tools analyze
applicant data—including résumé content, employment history, gaps in work, and other indicators—
to predict applicant suitability, including traits such as reliability, productivity, and expected
attendance.
57. 	These same data points—including interruptions in employment, medical-related
leave, or patterns consistent with treatment and recovery—are commonly correlated with underlying
health conditions such as cancer survivorship and respiratory impairments like asthma.
58. 	As recognized in the scholarly literature, algorithmic hiring tools can identify and rely
upon such proxy indicators of illness or health-related limitations—even where doing so would be
unlawful if performed directly—thereby enabling systems to infer the existence of a disability
without any explicit inquiry.
59. 	Because individuals with serious medical conditions, including cancer and asthma,
may exhibit distinct work histories, treatment-related gaps, or other behavioral and professional
patterns, these systems can disproportionately flag and screen out such applicants based on inferred
health status rather than job-related qualifications.
60. 	The companies listed above sent Ms. Hughes automated rejection notices for positions
for which she met or exceeded the posted hiring requirements.
61. 	In some instances, these automated rejection emails stated that Ms. Hughes “did not
meet the minimum requirements for the role,” despite her qualifications, experience, and credentials
demonstrating otherwise.
62. 	These rejections were frequently issued within minutes or a few hours of submission
and often occurred outside normal business hours—including times such as 2:40 a.m. on a Sunday
or 11:00 p.m. on a weekday—strongly indicating that her applications were screened and rejected by
automated systems rather than reviewed by any human decision-maker.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 18 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 18 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 20 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
19
PLAINTIFFS’ THIRD AMENDED COMPLAINT
63. 	This pattern of rapid, uniform rejection across multiple employers is consistent with
the operation of Workday’s algorithmic screening tools identifying and acting upon proxy indicators
of disability and health status—including those associated with cancer survivorship and asthma—
rather than conducting any individualized assessment of Ms. Hughes’ qualifications.
C. 	Sheilah Johnson-Rocha
64 	Sheilah Johnson-Rocha is an African-American female. 	She is over the age of 40.
Ms. Johnson-Rocha’s date of birth is January 30, 1983.
65, 	Ms. Johnson-Rocha has three college degrees: (1) an Associates in Arts (Business
Administration); Bachelor of Business of Administration; and, a Master of Science-Major in
Management. She also possesses various certifications within the pharmaceutical and biotechnology
industry. She has worked in the pharmaceutical and biotechnology industries for the past 21 years.
66. 	Since March 2023, Ms. Johnson-Rocha has submitted approximately 2,000 job
applications to various pharmaceutical and biotech companies.
67. 	A large majority of the companies to which she has applied use Workday’s job
application platform. Upon information and belief, her qualifications and experience exceeded all
criteria in the job requisitions for each of the roles for which she applied, however almost all of her
applications were met with an automated rejection notification.
68. 	Furthermore, many of the jobs for which she applied were not filled and are
continuously re-posted. When the jobs are re-posted, she re-applied only to receive an automated
rejection email.
69. 	In some cases, the automated rejection emails stated “You do not meet the minimum
requirements for this role” even though she met or exceeded all posted requirements based on her
professional resume. Automated rejections were often received within a few hours of applying or
were sent at odd times outside of business hours, indicating a human did not review the applications.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 19 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 19 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 21 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
20
PLAINTIFFS’ THIRD AMENDED COMPLAINT
D. 	Faithlinh Rowe
70. 	Plaintiff FaithLinh Rowe is an Asian American female over the age of forty (40). Ms.
Rowe was born in 1981 and is a member of protected classes based on race, sex, and age.
71. 	Ms. Rowe is a highly qualified Human Resources Information Systems (“HRIS”)
professional with extensive experience in human resources, including HR technology systems, talent
management, and enterprise HR operations. Her background includes roles requiring advanced
knowledge of HR platforms, data systems, and workforce management tools, making her well-
qualified for mid- to senior-level HR and HRIS positions.
72. 	Beginning as early as 2020 and continuing through 2025, Ms. Rowe applied for
numerous positions through hiring systems powered by Defendant Workday, Inc., including
positions for which she met or exceeded the stated qualifications. These applications included roles
across multiple employers and industries, as well as positions directly with Workday itself.
73. 	Between 2020 and April 2021, while residing in California, Ms. Rowe applied to at
least seven (7) positions associated with Defendant’s Pleasanton, California headquarters. Each of
these applications was submitted through Defendant’s centralized Workday platform and processed
through the same automated system:
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 20 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 20 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 22 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
21
PLAINTIFFS’ THIRD AMENDED COMPLAINT
74. 	Ms. Rowe received automated responses and rejections through the Workday
platform for these positions. The rejections were issued without any individualized feedback,
interview opportunity, or indication that her applications were meaningfully reviewed by a human
decisionmaker.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 21 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 21 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 23 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
22
PLAINTIFFS’ THIRD AMENDED COMPLAINT
75. 	Ms. Rowe has also applied for additional positions through Defendant’s Workday
platform, including positions for which she met or exceeded the stated qualifications, yet experienced
a consistent pattern of rejection.
76. 	On July 15, 2024, Ms. Rowe applied for an Associate Director – HR Technology
position at AT&T through the Workday platform. At the time of application, the position had been
posted for less than one hour and had only one identified applicant.
77. 	Immediately upon submitting her application, Ms. Rowe’s status was automatically
changed to “No longer under consideration,” without any interview, follow-up, or individualized
review.
78. 	The 	near-instantaneous 	rejection—occurring 	at 	the 	moment 	of 	submission—
demonstrates that her application was screened out by an automated system rather than a human
decisionmaker.
79. 	Following this rejection, Ms. Rowe contacted Jeff Ellis, Associate Director of Talent
Acquisition at AT&T, through LinkedIn to report the issue and request review of her application, but
received no response.
80. 	Ms. 	Rowe 	has 	retained 	documentary 	evidence 	of 	her 	applications, 	including
screenshots of the Workday application portal reflecting immediate rejection statuses, as well as
communications evidencing her attempts to seek review.
81. 	Across her applications submitted through Workday systems, Ms. Rowe experienced
a consistent pattern of rapid, uniform rejections without explanation, feedback, or any indication of
individualized human review.
82. 	Based on her professional experience working with HRIS platforms, including
Workday, Ms. Rowe understands that such immediate and uniform rejections are consistent with
automated screening, scoring, or ranking systems applied at or near the point of application
submission.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 22 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 22 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 24 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
23
PLAINTIFFS’ THIRD AMENDED COMPLAINT
83. 	Upon information and belief, the automated screening systems that evaluated Ms.
Rowe’s applications were designed, developed, maintained, and controlled by Defendant from its
headquarters in Pleasanton, California.
84. 	As a result of Defendant’s use of these automated screening systems, Ms. Rowe—an
Asian American woman over the age of forty—was repeatedly and systematically excluded from
employment opportunities despite her qualifications.
85. 	The uniformity and timing of these rejections are indicative of a centralized,
algorithmic screening system operating at scale, rather than individualized hiring decisions.
HOW ALGORITHMIC DISCRIMINATION WORKS
86. 	Workday designs, markets, and deploys algorithmic decision-making tools that power
applicant screening systems used by Workday and other employers to accept, rank, or reject job
applicants. These systems function as gatekeepers to employment opportunities and operate through
automated classification, scoring, and recommendation processes that materially affect hiring
outcomes. Defendant
87. 	Modern 	employment 	discrimination 	increasingly 	operates 	through 	data-driven
automation. Employers rely on algorithmic models to process large volumes of applications by
extracting, sorting, and evaluating data points such as keywords, prior employment, educational
background, and inferred behavioral traits. These processes—commonly referred to as “data mining”
and “data analytics”—replace individualized assessment with statistical pattern recognition.
88. 	However, data-driven systems “learn by example,” meaning that the outputs of such
systems are necessarily shaped by the data on which they are trained. As leading scholarship
recognizes, algorithms built on “inaccurate, biased, or unrepresentative data can…produce outcomes
biased along lines of race, sex, or other protected characteristics.”
89. 	For hiring purposes data is mined on the front-end from applications via an Applicant
Tracking System (“ATS), which can be located on the company’s website or extracted from
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 23 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 23 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 25 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
24
PLAINTIFFS’ THIRD AMENDED COMPLAINT
applicants on job boards. An applicant tracking system (ATS) is a software application that enables
the electronic handling of recruitment and hiring needs. Most job and resume boards (Reed Online,
LinkedIn.com, Monster.com, Hotjobs, CareerBuilder, Indeed.com) have partnerships with ATS
software providers to provide parsing support and easy data migration from one system to another.
90. 	In the hiring context, these systems are typically embedded within Applicant Tracking
Systems (“ATS”), which collect and process applicant data from employer portals and integrated job
boards. These systems parse resumes, extract features, and transform applicants into data profiles
that are subsequently scored or ranked through algorithmic models.
91. 	Contemporary ATS platforms, including those offered by Defendant, function as
integrated “platform-as-a-service” ecosystems, allowing third-party tools and machine-learning
models to seamlessly interact with applicant data. This architecture enables “omnichannel”
recruitment, in which algorithmic decision-making extends across job boards, employer sites, and
third-party databases, thereby amplifying the scale and impact of automated screening.
92. 	These 	systems 	rely 	on 	machine-learning 	algorithms 	that 	identify 	statistical
relationships within training data to predict applicant suitability. Critically, such systems do not
evaluate applicants based on merit in a vacuum; rather, they replicate patterns embedded in historical
data.
93. 	As courts and scholars increasingly recognize, the exclusion of explicit demographic
variables (such as race, age, or disability) does not render such systems neutral. Instead, algorithms
routinely rely on proxy variables—such as zip code, educational institution, employment gaps, or
language usage—that are highly correlated with protected characteristics. As a result, ostensibly
neutral systems can reproduce and even amplify discrimination.
94. 	Indeed, data mining systems may “inherit the prejudices of prior decision makers” or
“reflect the widespread biases that persist in society,” thereby producing systematically adverse
outcomes for protected groups.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 24 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 24 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 26 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
25
PLAINTIFFS’ THIRD AMENDED COMPLAINT
95. 	These risks are not hypothetical. In one widely reported example, Amazon abandoned
an automated hiring system after discovering that it systematically penalized female applicants.
Although the system did not include gender as an input, it learned to downgrade resumes containing
gender-associated indicators and to favor male-dominated career trajectories, demonstrating how
biased training data can produce discriminatory outcomes even absent explicit intent.
96. 	In such systems, the algorithm does not independently assess merit; rather, it identifies
statistical patterns within historical data and reproduces them. Where the underlying data reflects
historical exclusion or inequality, the resulting model will encode and perpetuate those disparities.
97. 	Upon information and belief, Defendant’s algorithmic decision-making tools operate
by learning from and optimizing for employer preferences, including historical hiring decisions. As
a result, if an employer’s past practices reflect bias—whether explicit or implicit—the algorithm will
internalize and replicate those preferences.
98. 	Upon information and belief, when Defendant’s systems detect that certain categories
of 	applicants 	are 	less 	frequently 	selected 	or 	advanced 	by 	employer 	clients, 	the 	system
correspondingly reduces the likelihood that such applicants will be recommended or surfaced in the
future. In this way, Defendant’s tools operationalize and scale discriminatory preferences.
99. 	Algorithmic decision-making systems are not inherently neutral. Rather, they are
subject to what scholars describe as “classification bias,” whereby automated systems sort
individuals in ways that systematically disadvantage protected groups, even in the absence of explicit
discriminatory intent.
100. 	Moreover, the technical literature confirms that fairness in algorithmic systems is
inherently constrained: it is often mathematically impossible to satisfy competing definitions of
fairness simultaneously, meaning that some groups will inevitably bear disproportionate error rates
or adverse outcomes.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 25 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 25 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 27 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
26
PLAINTIFFS’ THIRD AMENDED COMPLAINT
101. 	These dynamics are further compounded by the opacity of algorithmic systems. As
scholars have observed, algorithmic decision-making processes are often proprietary and resistant to
scrutiny, making it difficult to detect, explain, or remedy discriminatory effects once embedded in
the system.
102. 	Government agencies and regulators have likewise recognized the discriminatory
risks posed by algorithmic systems. For example, policymakers have warned that automated
decision-making can “reproduce existing patterns of discrimination” and “exacerbate existing
inequalities,” particularly when relying on historical or proxy data.
103. 	The absence of adequate safeguards—such as bias testing, transparency mechanisms,
or corrective constraints—creates a substantial risk that algorithmic systems will drift over time,
amplifying disparities as they continuously retrain on new data. This phenomenon, commonly
referred to as “AI drift,” can cause systems to deviate from their intended design and produce
increasingly discriminatory outcomes.
104. 	Upon information and belief, Defendant has failed to implement sufficient guardrails
to prevent such discriminatory outcomes, including failing to adequately audit, validate, or constrain
its algorithmic decision-making systems.
105. 	This 	failure 	reflects, 	at 	minimum, 	deliberate 	indifference 	to 	the 	foreseeable
discriminatory impact of Defendant’s technology and, upon information and belief, constitutes a
conscious decision to prioritize efficiency and scalability over compliance with federal anti-
discrimination laws.
106. 	The lack of diversity within the technology workforce responsible for designing and
training such systems further exacerbates these risks. As experts have noted, homogenous
development environments can embed structural bias into data models, leading to discriminatory
outputs that disproportionately harm historically marginalized groups.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 26 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 26 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 28 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
27
PLAINTIFFS’ THIRD AMENDED COMPLAINT
107. 	Accordingly, Defendant’s algorithmic decision-making tools do not merely reflect
neutral technological processes; they function as mechanisms that systematize, scale, and obscure
unlawful discrimination in employment.
108. 	Upon information and belief, Workday determines which candidates to recommend
based on the interests demonstrated by its client-employers in certain types of candidates, Workday
will offer recommendations that reflect whatever biases employers happen to exhibit.
109. 	Upon information and belief, if Workday’s algorithmic decision-making tools
observe that a client-employer disfavors certain candidates who are members of a protected class, it
will decrease the rate at which it recommends those candidates. Thus, the recommendation
algorithmic decision-making tool caters to the prejudicial preferences of the client-employer.
110. 	Algorithmic decision-making and data analytics are not, and should not be assumed
to be, race neutral, disability neutral, or age neutral. Too often, they reinforce and even exacerbate
historical and existing discrimination.
111. 	For example, a 2019 study found that a clinical algorithm that many hospitals were
using to determine which patients need care was biased: Black patients assigned the same level of
risk—and thus allocated the same health care resources—were much sicker than white patients. This
happened because the algorithm had been trained on historical health care spending data, which
reflects a history in which Black patients had less money to spend on their health care than white
patients. From this, the algorithm falsely concluded that Black patients were healthier than equally
sick white patients.
112. 	Upon information and belief, Workday’s artificial intelligence and machine learning
tools are capable of discerning or inferring an applicant’s physical and mental health status—even
where no such information is expressly provided—by analyzing behavioral, linguistic, and
employment-related data collected during the application process.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 27 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 27 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 29 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
28
PLAINTIFFS’ THIRD AMENDED COMPLAINT
113. 	These systems evaluate applicants using natural language processing, speech analysis,
and computer-based assessments, measuring features such as response timing, hesitation patterns,
communication style, and interaction behaviors, and comparing those inputs against training data
derived from prior candidates or incumbent employees.
114. 	In addition, these tools analyze résumé and application data reflecting employment
history—including gaps in work, leaves of absence, tenure patterns, and attendance-related
indicators—which are commonly correlated with underlying health conditions such as chronic
illness, physical impairments, and mental health disorders.
115. 	As recognized in the scholarly literature, algorithmic hiring systems may use such
information to predict traits like reliability, productivity, or attendance, including by identifying
proxies for illness or health-related limitations, even where doing so would be unlawful if performed
directly.
116. 	Because individuals with disabilities or health conditions may exhibit different speech
patterns, response speeds, work histories, or behavioral characteristics, these systems can use such
proxy variables to identify, infer, or predict the existence of a disability or medical condition without
any explicit inquiry.
117. 	Where, as here, the underlying training data reflects historical bias or underrepresents
individuals with disabilities, the resulting models systematically penalize these characteristics,
screening out applicants based on inferred health conditions rather than job-related qualifications.
118. 	In this way, Workday’s ostensibly neutral screening tools function as de facto medical
examinations or disability-related inquiries, as they are designed to elicit, detect, or act upon
information likely to reveal physical or mental impairments—such as illness-related absences, health
conditions, or limitations—thereby disproportionately excluding individuals with disabilities in
violation of the ADA.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 28 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 28 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 30 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
PLAINTIFFS’ THIRD AMENDED COMPLAINT
119. 	The Representative Plaintiffs’ repeated near-immediate rejections—often within
hours or minutes of submission—are consistent with the operation of these automated screening tools
identifying and acting upon such proxy indicators of disability and health status, rather than any
individualized assessment of his qualifications.
120. 	Workday 	expressly represents that 	its system can 	“automatically disposition”
candidates and move them through—or exclude them from—the hiring pipeline without human
intervention.
121. 	These tools operate as predictive classification systems that rank applicants based on
inferred likelihood of success, using patterns derived from historical workforce data.
122. 	Upon 	information 	and 	belief, 	such 	systems 	reproduce 	and 	amplify 	historical
disparities embedded in training data or reflected through proxy variables.
123. 	Because protected characteristics can be inferred from non-protected inputs (e.g.,
education, employment history, geography), algorithmic systems can replicate discrimination
without explicitly using protected traits.
124. 	Workday’s integration of tools such as pymetrics further embeds standardized
assessment-driven decision-making into its platform.
125. 	Upon information and belief, these tools are trained on incumbent employee data,
thereby reinforcing existing workforce demographics and perpetuating homogeneity.
126. 	This process “locks in” the status quo by favoring applicants who resemble prior hires,
disproportionately 	disadvantaging 	African-Americans, 	older 	workers, 	and 	individuals 	with
disabilities.
127 	Workday also facilitates employee referral and recommendation systems that further
entrench existing demographic patterns, amplifying disparate impact.
128. 	These practices constitute facially neutral policies that produce systemic exclusion of
protected groups through common mechanisms applied across all client employers.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 29 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 29 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 31 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
30
PLAINTIFFS’ THIRD AMENDED COMPLAINT
129. 	Detecting such discrimination requires examining outputs and disparities across
groups, as bias manifests in selection rates, exclusion patterns, and error rates rather than in facial
system design.
A. 	Agency and Control
130. 	Workday operates as more than a passive software vendor. Its platform is the
mandatory gateway through which applicants must pass to be considered for employment with client-
employers and. with Workday itself.
131. 	Employers delegate substantial aspects of their hiring function to Workday, including
applicant screening, ranking, and dispositioning.
132. 	Workday provides ongoing consulting, training, and system optimization services,
reflecting a continuous and integrated role in hiring decisions.
133. 	As a result, Workday acts as an agent of its client-employers and exercises control
over access to employment opportunities.
134. 	Workday’s algorithmic screening tools directly interfere with and determine whether
applicants—including Mobley and the putative class—can form employment relationships with
prospective employers.
B. 	Common Policy and Classwide Impact
135. 	Plaintiff challenges the use of Workday’s common algorithmic screening tools as a
uniform employment practice.
136. 	These tools operate through standardized mechanisms—including résumé parsing,
assessments, predictive scoring, and automated dispositioning—applied across all employers using
the platform.
137. 	While implementation details may vary slightly by employer, the core discriminatory
mechanisms are common and centrally designed by Workday.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 30 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 30 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 32 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
31
PLAINTIFFS’ THIRD AMENDED COMPLAINT
138. 	The effects of these practices are likewise common: systematic exclusion of African-
Americans, Asian-Americans, women, individuals over 40, and individuals with disabilities at the
initial screening stage.
139. 	The number of individuals affected by these practices is in the thousands, if not tens
of thousands.
C. 	Claims
140. 	Workday’s screening tools constitute employment practices that cause a disparate
impact on African-American, Asian-American, and female applicants in violation of Title VII, 42
U.S.C. § 2000e-2(k).
141. 	These practices are not job-related and are not consistent with business necessity.
142. 	Workday’s assessments and screening tools constitute prohibited disability-related
inquiries and/or medical examinations under the ADAAA and disparately impact individuals with
disabilities.
143. 	Workday’s screening tools also disparately impact applicants over the age of 40 in
violation of the ADEA.
COUNT ONE
Intentional Employment Discrimination in
Violation of Title VII of the Civil Rights Act of 1964
144. 	Plaintiffs recognize that the Court has previously dismissed Count One stated in ¶¶
86-143 of the first Amended Complaint and do not include such matters again except for the limited
purpose of preserving whatever rights they may have to appeal.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 31 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 31 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 33 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
32
PLAINTIFFS’ THIRD AMENDED COMPLAINT
COUNT TWO
Disparate Impact Discrimination on the Basis of Race and Sex in Violation of Title VII
of the Civil Rights Act of 1964 and Disability in Violation of the ADAAA of 2008
Class Representatives Derek L. Mobley, Jill E. Hughes, Sheilah Johnson-Rocha,
and Faithlih Rowe
145 	Plaintiffs restate and incorporate ¶¶86-143 above on the same basis as if those
allegations were restated verbatim as part of the allegations of this Count.
146. 	Workday commissioned an external auditor to conduct “bias audits” to determine
whether its job application platform and/or its AI screening process and features have disparate
impact on the basis of race or gender based on the results for 724,352 applicants for employment
with ten of its largest customers for the period ending in September 2024. Such Workday reported
data shows that Workday’s job application platform and AI screening process has disparate impact
against African-Americans at a statistically significant rate greater than what would be expected to
occur in the absence of such disparate impact or if Workday’s AI-powered screening and selection
process is race neutral.
147. 	That same data for 724,352 applicants for employment with ten of Workday’s largest
customers also shows that Workday’s job application platform and AI screening and selection
process has disparate impact against women. Workday has publicly announced that its “AI-powered
platform . . . automate[s] the talent acquisition process, enabling more effective candidate, screening,
ranking, and matching.”
147. 	It has also publicly announced and advertised that its “AI-powered” products,
Platform 	and 	services 	“streamlin[e] 	and 	expedit[e] 	hiring 	processes 	through 	automated
notifications ..., all in [customers] flow of work.”
148. 	It does that, in part, by integrating the HiredScore AI-powered recruitment and
scoring process into its Workday Recruiting platform to “attract, hire, and retain talent.”
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 32 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 32 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 34 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
33
PLAINTIFFS’ THIRD AMENDED COMPLAINT
149. 	Workday has also publicly represented that “[b]y integrating HiredScore’s advanced
AI capabilities with Workday’s human capital management system, organizations can automate
candidate screening,” that “[t]he combination of Workday Talent Management, Workday Skills
Cloud, and HiredScore’s Talent Orchestration 	solutions, . . . provide[s] customers with a
comprehensive . . . talent acquisition and internal mobility offering” that “better enable[s] recruiters
to use data to connect talent to open opportunities,” including identifying candidates whose skills
and experience most closely match a customer's open jobs.”
150. 	All of HiredScore’s AI-powered products and services were fully merged and
integrated into Workday Recruiting well before 2024.
151. 	Workday integrated HiredScore’s AI software with Workday’s applicant tracking
system (ATS), its Workday Recruiting platform and its human capital management (HCM) software
used for recruiting, hiring, promoting and retaining its own employees and also use by its customers
and subscribers to recruit, hire, promote and retain their employees. To do that, Workday and
HiredScore jointly established a pre-acquisition partnership that integrated their HR-related
technologies through a “bi-directional” integration with Workday Recruiting that: (1) provides
seamless data exchange and process coordination between Workday’s and HiredScore’s AI systems;
and (2) enables recruiters and hiring managers to use HiredScore’s AI candidate matching, screening,
and rediscovery software within Workday’s platform and system.
152. 	Since April 29, 2024. HiredScore has been an internal department of Workday.
Plaintiff Mobley applied for job opportunities that were subject to such integration of HiredScore’s
AI-powered recruitment and hiring process and products, including his applications for employment
opportunities with Uline on June 2, 2024, Cigna on August 21, 2024, G.E. Health Care on July 16,
2024, Cardinal Health Care in October 2024 and other Workday customer -employers.
153. 	The applications of other Plaintiffs who have opted-in to this case have been similarly
processed through that same integrated Workday process both before and since April 2024,
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 33 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 33 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 35 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
34
PLAINTIFFS’ THIRD AMENDED COMPLAINT
including, but not limited to, Plaintiffs Jill E. Hughes Sheliah Johnson-Rocha, Faithlihn Rowe and
others.
154. 	Workday’s job application platform and AI screening and selection features have
disparate impact against applicants 40 or more years old at a statistically significant rate of greater
than what would be expected in the absence of such disparate impact or if Workday’s AI-powered
screening and selection process were age neutral.
155. 	Representative Derek L. Mobley, an African-American male who is diagnosed with
anxiety and depression.
156. 	Representative Plaintiff Mobley’s diagnoses are covered by the ADAAA of 2008.
157. 	He is a cum laude graduate of Morehouse College with a bachelor’s degree in finance
and is an honors graduate of ITT Technical Institute, holding a Server+ certification.
158. 	Representative Plaintiff Mobley is a qualified professional who has applied for over
one hundred positions through Workday's platform and has been disproportionately affected by its
AI/ML products, which have a disparate impact on African-American and disabled applicants.
159. 	For example, Mobley has applied to Fiserv Solutions, LLC, Sedgwick Claims
Management Services, Inc., and ResMed, Inc.
160. 	He was rejected for every position despite meeting the experiential and educational
requirements. At least one rejection was received within one hour of his application.
161. 	Representative Plaintiff Sheilah Johnson-Rocha is an African-American female who
possesses three academic degrees (A.A., B.B.A., M.S. in Management), certifications, and 21 years
in pharma/biotech.
162. 	Representative Plaintiff Johnson-Rocha is a qualified professional who has submitted
~2,000 job applications through Workday’s platform and has been disproportionately affected by its
AI/ML products, which have a disparate impact on African-American applicants.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 34 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 34 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 36 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
35
PLAINTIFFS’ THIRD AMENDED COMPLAINT
163. 	Beginning in December 2023, Representative Plaintiff Johnson-Rocha submitted four
separate applications to Thermo-Fisher Scientific for positions for which she met the qualifications.
164. 	She received rejections for all these positions despite meeting the qualifications, and
they were sent at odd hours.
165. 	Representative Plaintiff Jill E. Hughes possesses a B.A. (Univ. of Cincinnati), M.A.
(Univ. of Chicago), 28+ years as Senior Medical Writer, Project Manager, Scientific Director.
166. 	She is a cancer survivor and suffers from asthma, which are both covered disabilities
under the ADAAA of 2008.
167. 	Representative Plaintiff Hughes is a qualified professional who has applied for
hundreds of positions through Workday’s platform and has been disproportionately affected by its
AI/ML products, which have a disparate impact on disabled applicants.
168. 	Representative Plaintiff Hughes submitted applications to the following companies
for which she met the qualifications: Ciena, Intel, Stryker, and Thermo-Fisher Scientific.
169. 	She received rejections for all these positions, which often cited unmet minimum
requirements despite her meeting the position’s qualifications, and these rejections were sent
at odd hours.
170. 	Plaintiff FaithLinh Rowe is an Asian American female over the age of forty (40). Ms.
Rowe was born in 1981 and is a member of protected class based on race and sex.
171. 	Ms. Rowe is a highly qualified Human Resources Information Systems (“HRIS”)
professional with extensive experience in human resources, including HR technology systems, talent
management, and enterprise HR operations. Her background includes roles requiring advanced
knowledge of HR platforms, data systems, and workforce management tools, making her well-
qualified for mid- to senior-level HR and HRIS positions.
172. 	Beginning as early as 2020 and continuing through 2025, Ms. Rowe applied for
numerous positions through hiring systems powered by Defendant Workday, Inc., including
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 35 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 35 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 37 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
36
PLAINTIFFS’ THIRD AMENDED COMPLAINT
positions for which she met or exceeded the stated qualifications. These applications included roles
across multiple employers and industries, as well as positions directly with Workday itself.
173. 	On July 15, 2024, Ms. Rowe applied for an Associate Director – HR Technology
position at AT&T through the Workday platform. At the time of application, the position had been
posted for less than one hour and had only one identified applicant.
174. 	Immediately upon submitting her application, Ms. Rowe’s status was automatically
changed to “No longer under consideration,” without any interview, follow-up, or individualized
review.
175. 	The 	near-instantaneous 	rejection—occurring 	at 	the 	moment 	of 	submission—
demonstrates that her application was screened out by an automated system rather than a human
decisionmaker.
176. 	Following this rejection, Ms. Rowe contacted Jeff Ellis, Associate Director of Talent
Acquisition at AT&T, through LinkedIn to report the issue and request review of her application but
received no response.
177. 	Ms. 	Rowe 	has 	retained 	documentary 	evidence 	of 	her 	applications, 	including
screenshots of the Workday application portal reflecting immediate rejection statuses, as well as
communications evidencing her attempts to seek review.
178. 	Across her applications submitted through Workday systems, Ms. Rowe experienced
a consistent pattern of rapid, uniform rejections without explanation, feedback, or any indication of
individualized human review.
179. 	Based on her professional experience working with HRIS platforms, including
Workday, Ms. Rowe understands that such immediate and uniform rejections are consistent with
automated screening, scoring, or ranking systems applied at or near the point of application
submission.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 36 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 36 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 38 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
37
PLAINTIFFS’ THIRD AMENDED COMPLAINT
180. 	The automated recruitment, hiring, promoting, retaining and otherwise screening and
scoring tools utilized by Workday for making selection decisions-to include hiring-discriminate on
the basis of race.
181. 	Employers have delegated hiring decisions to Workday who then utilize automated
recruitment, 	hiring, 	promoting, 	retaining, 	screening 	and 	scoring 	tools 	that 	consciously 	or
unconsciously discriminate against applicants on the basis of race. There is no business necessity
justifying the disparate impact these automated screening and scoring tools have on individuals in
this protected category. 	Workday’s AI screening and selection criteria and process are a pretext for
racial discrimination, as shown by the disparate impact of such practices which have no manifest
relationship to the position[s] in question and there are alternative employment practices with less
adverse impact and equal or greater utility that Workday and its customers have failed or refused to
develop or use.
182 	Workday’s employment practices operate as the functional equivalent of a pretext for
discrimination.
183. 	Because there are no guardrails to regulate Workday’s conduct, the screening tools it
utilizes provide a ready mechanism for discrimination.
184. 	Workday’s automated recruitment, hiring, promoting, retaining, screening and
scoring tools discriminated against the Representative Plaintiffs and the proposed class both within
and outside the liability period in this case.
185. 	As a direct result of Workday’s utilization of discriminatory automated recruitment,
hiring, promoting, retaining, screening and scoring tools as described above, the Representative
Plaintiffs and the class they seek to represent have suffered damages including, but not limited to,
lost past and future income, compensation, and benefits.
186. 	Workday has also interfered with the present and future employment prospects of
class members that have used its application platform in violation of Title VII.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 37 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 37 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 39 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
38
PLAINTIFFS’ THIRD AMENDED COMPLAINT
187. 	In the absence of a direct employment relationship, Workday can still be held liable
under Title VII for its discriminatory treatment of the class members because it has interfered with
their opportunity to gain employment.
COUNT THREE
Intentional Discrimination
Age Discrimination in Employment Act of 1967, 29 U.S.C. §§ 623(a)(1)
188. 	Plaintiffs recognize that the Court has previously dismissed Count One of the first
Amended Complaint and do not include such matters again except for the limited purpose of
preserving whatever rights they may have to appeal.
COUNT FOUR
Disparate Impact Discrimination Age Discrimination in Employment Act of 1967,
29 U.S.C. §§ 623(a)(2) Class Representatives Derek L. Mobley, Jill E. Hughes,
Sheilah Johnson-Rocha, and Faithlinh Rowe
189. 	Plaintiffs restate and incorporate ¶¶86-143 above on the same basis as if those
allegations were restated verbatim as part of the allegations of this Count Four.
190. 	Workday has publicly announced that its “AI-powered platform . . . automate[s] the
talent acquisition process, enabling more effective candidate, screening, ranking, and matching.”
191. 	It has also publicly announced and advertised that its “AI-powered” products,
Platform 	and 	services 	“streamlin[e] 	and 	expedit[e] 	hiring 	processes 	through 	automated
notifications ..., all in [customers] flow of work.”
192. 	It does that, in part, by integrating the HiredScore AI-powered recruitment and
scoring process into its Workday Recruiting platform to “attract, hire, and retain talent.”
193. 	Workday has also publicly represented that “[b]y integrating HiredScore’s advanced
AI capabilities with Workday’s human capital management system, organizations can automate
candidate screening,” that “[t]he combination of Workday Talent Management, Workday Skills
Cloud, and HiredScore’s Talent Orchestration 	solutions, . . . provide[s] customers with a
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 38 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 38 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 40 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
39
PLAINTIFFS’ THIRD AMENDED COMPLAINT
comprehensive . . . talent acquisition and internal mobility offering” that “better enable[s] recruiters
to use data to connect talent to open opportunities,” including identifying candidates whose skills
and experience most closely match a customer's open jobs.”
194. 	All of HiredScore’s AI-powered products and services were fully merged and
integrated into Workday Recruiting well before 2024.
195. 	Workday integrated HiredScore’s AI software with Workday’s applicant tracking
system (ATS), its Workday Recruiting platform and its human capital management (HCM) software
used for recruiting, hiring, promoting and retaining its own employees and also use by its customers
and subscribers to recruit, hire, promote and retain their employees. To do that, Workday and
HiredScore jointly established a pre-acquisition partnership that integrated their HR-related
technologies through a “bi-directional” integration with Workday Recruiting that: (1) provides
seamless data exchange and process coordination between Workday’s and HiredScore’s AI systems;
and (2) enables recruiters and hiring managers to use HiredScore’s AI candidate matching, screening,
and rediscovery software within Workday’s platform and system.
196. 	Since April 29, 2024. HiredScore has been an internal department of Workday.
Plaintiff Mobley applied for job opportunities that were subject to such integration of HiredScore’s
AI-powered recruitment and hiring process and products, including his applications for employment
opportunities with Uline on June 2, 2024, Cigna on August 21, 2024, G.E. Health Care on July 16,
2024, Cardinal Health Care in October 2024 and other Workday customer -employers.
197. 	The applications of other Plaintiffs who have opted-in to this case have been similarly
processed through that same integrated Workday process both before and since April 2024,
including, but not limited to, Plaintiffs Jille E. Hughes, Sheliah Johnson-Rocha, Faithlinh Rowe, and
others.
198. 	Workday’s job application platform and AI screening and selection features have
disparate impact against applicants 40 or more years old at a statistically significant rate of greater
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 39 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 39 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 41 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
40
PLAINTIFFS’ THIRD AMENDED COMPLAINT
than what would be expected in the absence of such disparate impact or if Workday’s AI-powered
screening and selection process were age neutral.
199. 	Representative Derek L. Mobley, an African-American male born in 1974, is over 40
and manages anxiety and depression.
200. 	He graduated cum laude from Morehouse College in 1995 with a bachelor’s degree
in finance and is an honors graduate of ITT Technical Institute, holding a Server+ certification.
201. 	Representative Plaintiff Mobley is a qualified professional who has applied for over
one hundred positions through Workday's platform and has been disproportionately affected by its
AI/ML products, which have a disparate impact on applicants over the age of 40.
202. 	For example, Mobley has applied to Fiserv Solutions, LLC, Sedgwick Claims
Management Services, Inc., and ResMed, Inc.
203. 	He was rejected for every position despite meeting the experiential and educational
requirements. At least one rejection was received within one hour of his application.
204. 	Representative Plaintiff Jill E. Hughes is a white female who possesses a B.A. (Univ.
of Cincinnati), M.A. (Univ. of Chicago), 28+ years as Senior Medical Writer, Project Manager,
Scientific Director. She was born in 1974 and is over the age of 40.
205. 	Representative Plaintiff Hughes is a qualified professional who has applied for
hundreds of positions through Workday's platform and has been disproportionately affected by its
AI/ML products, which have a disparate impact on applicants over the age of 40.
206. 	Representative Plaintiff Hughes submitted applications to the following companies
for which she met the qualifications: Ciena, Intel, Stryker, and Thermo-Fisher Scientific.
207. 	She received rejections for all these positions, which often cited unmet minimum
requirements despite her meeting the position’s qualifications, and these rejections were sent at odd
hours.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 40 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 40 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 42 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
41
PLAINTIFFS’ THIRD AMENDED COMPLAINT
208. 	Representative Plaintiff Sheilah Johnson-Rocha is an African-American female who
possesses three academic degrees (A.A., B.B.A., M.S. in Management), certifications, and 21 years
in pharma/biotech. She was born in 1983 and is over the age of 40.
209. 	Representative Plaintiff Johnson-Rocha is a qualified professional who has submitted
~2,000 job applications through Workday's platform and has been disproportionately affected by its
AI/ML products, which have a disparate impact on applicants over the age of 40.
210. 	Beginning in December 2023, Representative Plaintiff Johnson-Rocha submitted four
separate applications to Thermo-Fisher Scientific for positions for which she met the qualifications.
211. 	She received rejections for all these positions despite meeting the qualifications, and
they were sent at odd hours.
212. 	Representative Plaintiff Faithlinh Rowe is an Asian-American female with many
years of , Human Resource experience. She was born in 1981 and is over the age of 40.
213. 	Representative Plaintiff Rowe is a qualified professional who has submitted numerous
applications Workday's platform and has been disproportionately affected by its AI/ML products,
which have a disparate impact on applicants over the age of 40.
214. 	Representative Plaintiff Rowe applied for an Associate Director HR Technology
role at AT&T on Jul 15, 2024, and received an immediate rejection, despite meeting the position’s
qualifications.
215. 	Likewise, on May 12, 2025, at 1:10 p.m., Representative Plaintiff Rowe applied for a
Director, Human Resource Information Systems position with Kyndryl, Inc., by 1:40 p.m. (30
minutes later) she received a rejection email from kyndryl@myworkday.com.
216. 	The automated recruitment, hiring, promoting, retaining and otherwise screening and
scoring tools utilized by Workday for making selection decisions-to include hiring-discriminate on
the basis of age.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 41 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 41 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 43 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
42
PLAINTIFFS’ THIRD AMENDED COMPLAINT
217. 	Workday maintains discriminatory policies, patterns, and/or practices that have an
adverse impact on employees ages 40 and older in violation of the ADEA and are not, and cannot
be, justified by reasonable factors other than age.
218. 	Employers have delegated hiring decisions to Workday who then, upon information
and 	belief, 	utilize 	discriminatory 	algorithmic 	decision-making 	tools 	that 	consciously 	or
unconsciously discriminate against applicants on the basis of age.
219. 	For purposes of the ADEA, Workday is also an agent and/or indirect employer
because (1) it has been delegated authority to make hiring decisions by direct employers and (2) it
has the ability to interfere with and control access to employment opportunities with direct
employers.
220. 	There is no business necessity justifying the disparate impact these automated
recruitment, hiring, promoting, retaining and otherwise screening and scoring tools have on
individuals in this protected category.
221. 	Workday’s AI screening and selection criteria and process are a pretext for age
discrimination, as shown by the disparate impact of such practices which have no manifest
relationship to the position[s] in question and there are alternative employment practices with less
adverse impact and equal or greater utility that Workday and its customers have failed or refused to
develop or use.
222. 	Workday’s employment practices operate as the functional equivalent of a pretext for
age discrimination. Workday used discriminatory automated 	recruitment, hiring, promoting,
retaining and otherwise screening and scoring tools both within and outside the liability period in
this case.
223. 	As a direct result of Workday’s discriminatory policies and/or practices as described
above, the Representative Plaintiff and the collective he seeks to represent have suffered damages
including, but not limited to, lost past and future income, compensation, and benefits.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 42 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 42 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 44 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
43
PLAINTIFFS’ THIRD AMENDED COMPLAINT
COUNT FIVE
Intentional Discrimination
42 U.S.C. § 1981
224. 	Plaintiffs recognize that the Court has previously dismissed Count Five stated in the
first Amended Complaint and do not include such matters again except for the limited purpose of
preserving whatever rights they may have to appeal.
COUNT SIX
Aiding and Abetting Race, Disability, and Age Discrimination
Cal. Gov. Code §12940(I)
225. 	Plaintiffs recognize that the Court has previously dismissed Count Six stated in the
first Amended Complaint and do not include such matters again except for the limited purpose of
preserving whatever rights they may have to appeal.
COUNT SEVEN
Disparate Impact Discrimination on the
Basis of Gender in Violation Title VII of the Civil Rights Act of 1964
and California Fair Employment and Housing Act (FEHA)
Gov. Code § 12940 et seq.
226. 	Plaintiffs restate and incorporate ¶¶86-143 above on the same basis as if those
allegations were restated verbatim as part of the allegations of this Count Four.
227. 	Defendant’s applicant-screening systems—including its artificial intelligence and
machine-learning tools used to evaluate, rank, and reject applicants—are designed, developed,
maintained, and controlled from its California headquarters. Workday maintains its headquarters and
principal place of business at 6110 Stoneridge Mall Road, Pleasanton, California.
228. 	Evidence shows that Workday has trained its artificial intelligence procedures to
screen and score applicants in a manner that has exhibited “impermissible bias” based on either race,
gender, age or disability.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 43 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 43 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 45 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
44
PLAINTIFFS’ THIRD AMENDED COMPLAINT
229. 	Workday trained its AI models and algorithmic tools on the basis of its own
proprietary datasets which contain over 200 million people profiles and customer data contributions,
and which also leveraged customer-contributed data to form test datasets consisting of thousands of
real-world job requisitions and millions of job applications. That same training dataset used to train
Workday’s AI algorithms and models is common to all of Workday’s customers and subscribers
worldwide, many of whom are located in California or otherwise serviced by Workday from
California.
230. 	Such algorithmic training and dataset have been designed, developed, operated from
Workday’s California operations and facilities, not only to screen score and rank Workday’s own
applicants for jobs in California, but to simultaneously train and operate those same AI applicant
screening services Workday provides to its out-of-state customers and subscribers from California.
231. 	Workday specifically designed and trained all such AI models, algorithms and
application screening process to copy or mimic past patterns of hiring decisions, behaviors and
selection criteria. It has done so even though it knew (and admits) that such training process and data
process has an inherent susceptibility to carrying forward past impermissible bias into the present
and has a significant likelihood that the algorithmic tool's output will contain such impermissible
bias based on either race, gender, age or disability.
232. 	As a result, the discriminatory conduct at issue—including applicant screening,
scoring, and rejection—originates in and is carried out from California, causing injury to applicants
who are denied employment opportunities, wages, and associated benefits, including backpay.
233. 	Workday has also failed to implement adequate bias mitigation, auditing, or
safeguards as required under California law, including the FEHA, thereby perpetuating systemic
disparities through its centralized, California-based hiring infrastructure.
234. 	For example, between 2020 and April 2021, while residing in California, Plaintiff
FaithLinh Rowe applied for at least seven (7) positions tied to or at Defendant’s Pleasanton,
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 44 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 44 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 46 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
45
PLAINTIFFS’ THIRD AMENDED COMPLAINT
California headquarters. Each of these applications was submitted through and processed by
Defendant’s automated application platform and screening procedure for all Workday customers and
Workday itself.
237. 	Those application records reflect that Plaintiff Rowe’s submissions were evaluated
and rejected through that same Workday hiring applicant platform for processing applicant data and
generating hiring outcomes for all Workday customers and Workday itself.
238. 	Ms. Rowe also received automated responses and rejections from Workday’s
platform. Her rejection notices were issued without any individualized feedback, interview
opportunity, or indication that her applications were meaningfully reviewed by machine or human
decisionmaker.
239. 	Based on her professional experience working with HRIS platforms, including
Workday, Ms. Rowe recognized that this pattern of immediate and uniform rejections are indicative
of, and consistent with an automated algorithmic screening, scoring, or ranking system applied at or
near the point of applying, rather than individualized non-mechanized screening and decisionmaking.
240. 	The automated screening systems that evaluated Ms. Rowe’s applications were
designed, developed, maintained, and controlled by Defendant from its headquarters in Pleasanton,
California.
241. 	The discriminatory conduct alleged herein—including the design, operation, and
application of Defendant’s automated screening systems—occurred in, was directed from, and is
substantially connected to this District.
242. 	As a result of Defendant’s use of these automated screening systems, Ms. Rowe—an
Asian American woman over the age of forty—was repeatedly and systematically excluded from
employment opportunities despite her qualifications.
243. 	In addition to Representative Plaintiff FaithLinh Rowe, numerous other individuals
applied for Workday jobs located in California using the same Workday-hosted application platform
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 45 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 45 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 47 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
46
PLAINTIFFS’ THIRD AMENDED COMPLAINT
systems that they also used to apply for jobs with other employers and that operates as the exclusive
gateway for employment opportunities in both California and throughout the United States.
244. 	Other examples of such applicants who applied for jobs in California include, but are
not limited to, the following Opt-In In Plaintiffs: Jason Adams, Diana El-Berry, Anthony Fox,
Cynthia Monfredi, Katrina Nelson, Roger Mosley, David Stephenson, Denise Russell, Elizabeth
Louis, Kenyon Johnston, Michael Street, Tracey Harriott, Theresa Gal, Shanice Love, Leigh-Ann
Carnegie, Stephen Duclos, John Erwin, Jr., Felicia Ives, Dionne Joseph, Noah Westwind, Tunde
Adedoyin, Steve Hurry, and Christopher Phelan. See, ECF. 220-1.
245. 	Evidence 	shows 	that 	these 	applications 	were 	processed 	through 	Workday’s
centralized application platform, where recruitment, application screening, AI-driven scoring, and
hiring selection decisions are designed, developed, and executed from Workday’s headquarters in
Pleasanton, California. Workday’s algorithmic systems—developed, trained, and governed in
California—function as the effective decisionmakers by evaluating applicants, assigning scores,
ranking candidates, and effectuating automated rejection decisions.
246. 	Representative Plaintiff Jill E. Hughes is a white female who possesses a B.A. (Univ.
of Cincinnati), M.A. (Univ. of Chicago), 28+ years as Senior Medical Writer, Project Manager,
Scientific Director.
247. 	Representative Plaintiff Hughes is a qualified professional who has applied for
hundreds of positions through Workday's platform and has been disproportionately affected by its
AI/ML products, which have a disparate impact on female applicants.
248. 	Representative Plaintiff Hughes submitted applications to the following companies
for which she met the qualifications: Ciena, Intel, Stryker, and Thermo-Fisher Scientific.
249. 	She received rejections for all these positions, which often cited unmet minimum
requirements despite her meeting the position’s qualifications, and these rejections were sent at odd
hours.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 46 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 46 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 48 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
47
PLAINTIFFS’ THIRD AMENDED COMPLAINT
250. 	Representative Plaintiff Sheilah Johnson-Rocha is an African-American female who
possesses three academic degrees (A.A., B.B.A., M.S. in Management), certifications, and 21 years
in pharma/biotech.
251. 	Representative Plaintiff Johnson-Rocha is a qualified professional who has submitted
2,000 job applications through Workday's platform and has been disproportionately affected by its
AI/ML products, which have a disparate impact on female applicants.
252. 	Beginning in December 2023, Representative Plaintiff Johnson-Rocha submitted four
separate applications to Thermo-Fisher Scientific for positions for which she met the qualifications.
253. 	She received rejections for all these positions despite meeting the qualifications, and
they were sent at odd hours.
254. 	Each application was subjected to the same automated screening, evaluation, and
dispositioning processes originating from Defendant’s California-based operations, and resulted in
her rejection.
255. 	The automated recruitment, hiring, promoting, retaining and otherwise screening and
scoring tools utilized by Workday for making selection decisions-to include hiring-discriminate on
the basis of gender.
256. 	Workday commissioned an external auditor to conduct “bias audits” to determine
whether its job application platform and/or its AI screening process and features have disparate
impact on the basis of gender based on the results for 724,352 applicants for employment with ten of
its largest customers for the period ending in September 2024.
257. 	That data for 724,352 applicants for employment with ten of Workday’s largest
customers shows that Workday’s job application platform and AI screening and selection process
has disparate impact against women greater than what would be expected to occur in the absence of
such disparate impact or if Workday’s AI-powered screening and selection process were gender
neutral.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 47 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 47 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 49 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
48
PLAINTIFFS’ THIRD AMENDED COMPLAINT
258. 	Workday has publicly announced that its “AI-powered platform . . . automate[s] the
talent acquisition process, enabling more effective candidate, screening, ranking, and matching.”
259. 	It has also publicly announced and advertised that its “AI-powered” products,
Platform 	and 	services 	“streamlin[e] 	and 	expedit[e] 	hiring 	processes 	through 	automated
notifications ..., all in [customers] flow of work.”
260. 	It does that, in part, by integrating the HiredScore AI-powered recruitment and
scoring process into its Workday Recruiting platform to “attract, hire, and retain talent.”
261. 	Workday has also publicly represented that “[b]y integrating HiredScore’s advanced
AI capabilities with Workday’s human capital management system, organizations can automate
candidate screening,” that “[t]he combination of Workday Talent Management, Workday Skills
Cloud, and HiredScore’s Talent Orchestration 	solutions, . . . provide[s] customers with a
comprehensive . . . talent acquisition and internal mobility offering” that “better enable[s] recruiters
to use data to connect talent to open opportunities,” including identifying candidates whose skills
and experience most closely match a customer's open jobs.”
262. 	All of HiredScore’s AI-powered products and services were fully merged and
integrated into Workday Recruiting well before 2024.
263. 	Workday integrated HiredScore’s AI software with Workday’s applicant tracking
system (ATS), its Workday Recruiting platform and its human capital management (HCM) software
used for recruiting, hiring, promoting and retaining its own employees and also use by its customers
and subscribers to recruit, hire, promote and retain their employees. To do that, Workday and
HiredScore jointly established a pre-acquisition partnership that integrated their HR-related
technologies through a “bi-directional” integration with Workday Recruiting that: (1) provides
seamless data exchange and process coordination between Workday’s and HiredScore’s AI systems;
and (2) enables recruiters and hiring managers to use HiredScore’s AI candidate matching, screening,
and rediscovery software within Workday’s platform and system.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 48 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 48 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 50 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
49
PLAINTIFFS’ THIRD AMENDED COMPLAINT
264. 	Employers have delegated hiring decisions to Workday who then utilize automated
recruitment, 	hiring, 	promoting, 	retaining, 	screening 	and 	scoring 	tools 	that 	consciously 	or
unconsciously discriminate against applicants on the basis of gender. There is no business necessity
justifying the disparate impact these automated screening and scoring tools have on individuals in
this protected category.
265. 	Workday’s AI screening and selection criteria and process are a pretext for racial
discrimination, as shown by the disparate impact of such practices which have no manifest
relationship to the position[s] in question and there are alternative employment practices with less
adverse impact and equal or greater utility that Workday and its customers have failed or refused to
develop or use.
266. 	Because there are no guardrails to regulate Workday’s conduct, the screening tools it
utilizes provide a ready mechanism for discrimination.
267. 	Workday’s automated recruitment, hiring, promoting, retaining, screening and
scoring tools discriminated against the Representative Plaintiffs and the proposed class both within
and outside the liability period in this case.
268. 	As a direct result of Workday’s utilization of discriminatory automated recruitment,
hiring, promoting, retaining, screening and scoring tools as described above, the Representative
Plaintiffs and the class they seek to represent have suffered damages including, but not limited to,
lost past and future income, compensation, and benefits.
269. 	Workday has also interfered with the present and future employment prospects of
class members that have used its application platform in violation of Title VII and California Fair
Employment and Housing Act (FEHA) Gov. Code § 12940 et seq.
270. 	In the absence of a direct employment relationship Workday can still be held liable
under Title VII and FEHA for its discriminatory treatment of the class members because it has
interfered with their opportunity to gain employment.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 49 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 49 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 51 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
50
PLAINTIFFS’ THIRD AMENDED COMPLAINT
COUNT EIGHT
Disparate Impact Discrimination on the Basis of Race,
Gender, and Age in violation of the
California Fair Employment and Housing Act (FEHA)
Gov. Code § 12940 et seq.
271. 	Plaintiffs restate and incorporate ¶¶86-143, 227-270 above on the same basis as if
those allegations were restated verbatim as part of the allegations of this Count Four.
272. 	Workday commissioned an external auditor to conduct “bias audits” to determine
whether its job application platform and/or its AI screening process and features have disparate
impact on the basis of race or gender for 724,352 applicants for employment with ten of its largest
customers for the period ending in September 2024.
273. 	Such Workday-reported data shows that Workday’s job application platform and AI
screening process has disparate impact against African-Americans at a statistically significant rate
greater than what would be expected to occur in the absence of such disparate impact or if Workday’s
AI-powered screening and selection process is race neutral.
274. 	That same data for 724,352 applicants for employment with ten of Workday’s largest
customers also shows that Workday’s job application platform and AI screening and selection
process has disparate impact against women. Workday has publicly announced that its “AI-powered
platform . . . automate[s] the talent acquisition process, enabling more effective candidate, screening,
ranking, and matching.”
275. 	Workday has publicly announced that its “AI-powered platform . . . automate[s] the
talent acquisition process, enabling more effective candidate, screening, ranking, and matching.” It
has also publicly announced and advertised that its “AI-powered” products, Platform and services
“streamlin[e] and expedit[e] hiring processes through automated notifications ..., all in [customers]
flow of work.”
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 50 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 50 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 52 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
51
PLAINTIFFS’ THIRD AMENDED COMPLAINT
276. 	It does that, in part, by integrating the HiredScore AI-powered recruitment and
scoring process into its Workday Recruiting platform to “attract, hire, and retain talent.”
277. 	Workday has also publicly represented that “[b]y integrating HiredScore’s advanced
AI capabilities with Workday’s human capital management system, organizations can automate
candidate screening,” that “[t]he combination of Workday Talent Management, Workday Skills
Cloud, and HiredScore’s Talent Orchestration 	solutions, . . . provide[s] customers with a
comprehensive . . . talent acquisition and internal mobility offering” that “better enable[s] recruiters
to use data to connect talent to open opportunities,” including identifying candidates whose skills
and experience most closely match a customer's open jobs.”
278. 	All of HiredScore’s AI-powered products and services were fully merged and
integrated into Workday Recruiting well before 2024.
279. 	Workday integrated HiredScore’s AI software with Workday’s applicant tracking
system (ATS), its Workday Recruiting platform and its human capital management (HCM) software
used for recruiting, hiring, promoting and retaining its own employees and also use by its customers
and subscribers to recruit, hire, promote and retain their employees. To do that, Workday and
HiredScore jointly established a pre-acquisition partnership that integrated their HR-related
technologies through a “bi-directional” integration with Workday Recruiting that: (1) provides
seamless data exchange and process coordination between Workday’s and HiredScore’s AI systems;
and (2) enables recruiters and hiring managers to use HiredScore’s AI candidate matching, screening,
and rediscovery software within Workday’s platform and system.
280. 	Since April 29, 2024, HiredScore has been an internal department of Workday.
Plaintiff Mobley applied for job opportunities that were subject to such integration of HiredScore’s
AI-powered recruitment and hiring process and products, including his applications for employment
opportunities with Uline on June 2, 2024, Cigna on August 21, 2024, G.E. Health Care on July 16,
2024, Cardinal Health Care in October 2024 and other Workday customer-employers.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 51 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 51 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 53 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
52
PLAINTIFFS’ THIRD AMENDED COMPLAINT
281. 	The applications of other Plaintiffs who have opted-in to this case have been similarly
processed through that same integrated Workday process both before and since April 2024,
including, but not limited to, Plaintiffs Jill E. Hughes, Sheliah Johnson-Rocha, Faithlinh Rowe and
others.
282. 	Workday’s job application platform and AI screening and selection features have
disparate impact against applicants 40 or more years old at a statistically significant rate of greater
than three standard deviations from what would be expected in the absence of such disparate impact
or if Workday’s AI-powered screening and selection process were age neutral.
283. 	Representative Plaintiff Rowe is an Asian-American female over the age of 40, with
many years of Human Resource experience.
284. 	Beginning as early as 2020 and continuing through 2025, Representative Plaintiff
FaithLinh Rowe applied for numerous positions through hiring systems powered by Defendant
Workday, Inc., including positions for which she met or exceeded the stated qualifications.
285. 	These applications spanned multiple employers and industries and included positions
with Workday itself, all of which were submitted through Defendant’s centralized platform and
processed using its automated screening systems.
286. 	Between 2020 and April 2021, while residing in California, Ms. Rowe applied to at
least seven (7) positions associated with Defendant’s Pleasanton, California headquarters.
287. 	Each application was subjected to the same automated screening, evaluation, and
dispositioning processes originating from Defendant’s California-based operations and resulted in
her rejection.
288. 	As a result of these practices, Ms. Rowe was disproportionately impacted by
Defendant’s AI- and machine learning-driven screening tools, which have a disparate impact on
female applicants.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 52 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 52 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 54 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
53
PLAINTIFFS’ THIRD AMENDED COMPLAINT
289. 	Representative Plaintiff Rowe applied for an Associate Director HR Technology role
at AT&T on Jul 15, 2024, and received an immediate rejection, despite meeting the position’s
qualifications. Likewise, on May 12, 2025, at 1:10 p.m., Representative Plaintiff Rowe applied for a
Director, Human Resource Information Systems position with Kyndryl, Inc., by 1:40 p.m. (30
minutes later) she received a rejection email from kyndryl@myworkday.com.
290. 	The automated recruitment, hiring, promoting, retaining and otherwise screening and
scoring tools utilized by Workday for making selection decisions-to include hiring-discriminate on
the basis of race and age.
291. 	Employers have delegated hiring decisions to Workday who then utilize automated
recruitment, 	hiring, 	promoting, 	retaining, 	screening 	and 	scoring 	tools 	that 	consciously 	or
unconsciously discriminate against applicants on the basis of race and age. 	There is no business
necessity justifying the disparate impact these automated screening and scoring tools have on
individuals in this protected category.
292. 	Workday’s AI screening and selection criteria and process are a pretext for racial
discrimination, as shown by the disparate impact of such practices which have no manifest
relationship to the position[s] in question and there are alternative employment practices with less
adverse impact and equal or greater utility that Workday and its customers have failed or refused to
develop or use. Workday’s employment practices operate as the functional equivalent of a pretext
for discrimination.
293. 	Because there are no guardrails to regulate Workday’s conduct, the screening tools it
utilizes provide a ready mechanism for discrimination.
294. 	Workday’s automated recruitment, hiring, promoting, retaining, screening and
scoring tools discriminated against Representative Plaintiff Rowe and the proposed class both within
and outside the liability period in this case.
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 53 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 53 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 55 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
54
PLAINTIFFS’ THIRD AMENDED COMPLAINT
295. 	As a direct result of Workday’s utilization of discriminatory automated recruitment,
hiring, promoting, retaining, screening and scoring tools as described above, Representative Plaintiff
Rowe and the class she seeks to represent have suffered damages including, but not limited to, lost
past and future income, compensation, and benefits.
296. 	Workday has also interfered with the present and future employment prospects of
class members that have used its application platform in violation of California Fair Employment
and Housing Act (FEHA) Gov. Code § 12940 et seq.
297. 	In the absence of a direct employment relationship Workday can still be held liable
under FEHA for its discriminatory treatment of the class members because it has interfered with their
opportunity to gain employment.
PRAYER FOR RELIEF
WHEREFORE, the Representative Plaintiff and the Proposed Classes pray for relief as
follow:
1. 	Certification of the case as a class action on behalf the proposed subclasses;
2. 	Designation of Plaintiff as representative of the subclasses;
3. 	Designation of Plaintiff’s Counsel of record as Class Counsel;
4. 	A declaratory judgment that the practices complained of herein are unlawful and
violate Title VII, the ADEA, the ADAAA, and Cal. Gov. Code §12940(I);
5. 	A preliminary and permanent injunction against the Company and its officers, agent,
successors employees, representatives, and any and all persons acting in correct with them from
engaging in each of the unlawful policies, practices, customs, and usages set forth herein;
6. 	An order that the Company institute and carry out policies, practices, and programs
that provide equal employment opportunities for all minorities, and that it eradicate the effects of its
past and present unlawful employment practices;
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 54 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 54 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 56 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
55
PLAINTIFFS’ THIRD AMENDED COMPLAINT
7. 	For back pay, front pay and other monetary relief according to proof (including
interest and benefits);
8. 	For all damages sustained as a result of the Company’s conduct according to proof;
9. 	For compensatory damages, nominal damages, and liquidated damages according to
proof;
10. 	For exemplary damages in an amount commensurate with the Company’s ability to
pay, to deter future conduct, and to set an example for others;
11. 	For reasonable attorneys’ fees and costs including under to the extent allowable by
law;
12. 	Pre-judgment and post-judgment interest, as provided by law;
13. 	For such ancillary orders, decrees and such further legal and equitable relief as may
be necessary to enjoin and restrain the improper conduct and wrongdoing of Defendant; and
14. 	For such other and further relief as the Court deems proper.
JURY TRIAL DEMANDED
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 55 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 55 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 57 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
56
PLAINTIFFS’ THIRD AMENDED COMPLAINT
/s/Roderick T. Cooks 	/s/Robert L. Wiggins, Jr.
Lee D. Winston (ASB: 6407O72L) 	Robert L. Wiggins, Jr. (ASB: 1754G63R)
Admitted pro hac vice 	Admitted Pro Hac Vice
lwinston@winstoncooks.com 	rwiggins@wigginschilds.com
Roderick T. Cooks (ASB: 5819O78R) 	Ann K. Wiggins (ASB: 7006i61a)
Admitted Pro Hac Vice 	Admitted Pro Hac Vice
rcooks@winstoncooks.com 	awiggins@wigginschilds.com
Bethany Mae Logan (ASB: 1360X15Z) 	Samuel Fisher (ASB: 2675E68S)
Admitted Pro Hac Vice 	Admitted Pro Hac Vice
bneal@winstoncooks.com 	sf@wgginschilds.com
Winston Cooks, LLC 	Jennifer Wiggins Smith (ASB:9622E53S)
420 20th Street North, Suite 2200 	Admitted Pro Hac Vice
Birmingham, Alabama 35203 	jsmith@wigginschilds.com
Telephone: (205) 502-0970 	Nicki L. Lawsen (ASB: 2602C00K)
Facsimile: (205) 278-5876 	Admitted Pro Hac Vice
nlawsen@wigginschilds.com
Freddrick M. Moore (ASB: 7952L20o
mmoore@wiggins childs.com
Admitted Pro Hac Vice
Wiggins Childs Pantazis Fisher and
Goldfarb
301 North 19th Street
Birmingham, Alabama 35203
Telephone: (205) 314-0500
Facsimile: 205) 254-1500
Local Counsel:
Jay Greene, JD, CPA
greenattorney@gmail.com
Greene Estate, Probate, and Elder Law Firm
447 Sutter Street, Suite 435
San Francisco, CA 94108
Telephone: (415) 905-0215
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 56 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 56 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 58 of 59

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
57
PLAINTIFFS’ THIRD AMENDED COMPLAINT
CERTIFICATE OF SERVICE
I hereby certify that on March 27, 2026. I electronically filed the foregoing document with
the United States District Court for the Northern District of California by using the CM/ECF system.
I certify that the following parties or their counsel of record are registered as ECF Filers and that they
will be served by the CM/ECF system:
Julie Ann Totten 	Jtotten@orrick.com, jponce@orrick.com
Erin M. Connell 	econnell@orrick.com
Kyla Delgado Grundy kgrundy@ orrick.com
Alexandria R. Elliott (aelliott@orrick.com)
s/Roderick T. Cooks
Of Counsel
Case 3:23-cv-00770-RFL 	Document 287 	Filed 03/27/26 	Page 57 of 57	Case 3:23-cv-00770-RFL 	Document 369-1 	Filed 06/29/26 	Page 57 of 57	Case 3:23-cv-00770-RFL Document 372 Filed 07/01/26 Page 59 of 59