Is AI résumé screening actually illegal?

Yes, under certain conditions. The US federal court in Mobley v. Workday (2024–2025) permitted claims that AI screening vendors can be held directly liable as "agents" under Title VII (race/gender/religion), the ADEA (age), and the ADA (disability). The EEOC has listed AI in hiring as a strategic enforcement priority for 2024–2028. However, employers and vendors can defend themselves if they audited for bias, documented the audit, and remediated known gaps — so bias itself is not automatically illegal; negligence and indifference are.

How biased is the University of Washington study?

The 2024 UW / AIES study (Wilson & Caliskan, NIST-funded, peer-reviewed) tested 120 large language models against 554 résumés and 500 job descriptions — 3 million+ comparisons total. It found white-associated names preferred 85% of the time vs 9% for Black-associated names. Male names preferred 52% vs female 11%. Black-female intersectional names never preferred over white-male. The methodology is sound (LLM ranking under realistic job-matching prompts) and reproducible. The caveat: the study does not claim 100% of vendors are biased or that all LLMs are equally biased — it documents that bias *exists* at scale across mainstream models. Vendors such as Workday and HireVue tested since then show mixed results depending on their specific implementation and data curation.

What have AI hiring vendors said in response?

Workday and HireVue have published limited public statements. Workday announced bias-detection audits and committed to transparency (2024), but did not release results. HireVue phased out facial-analysis tools (2021, in response to earlier bias research) and pivoted to text-based screening, but full bias audits remain internal. Neither has released comprehensive third-party bias reports. ResumeBuilder's 2024 survey found 67% of business leaders believe AI hiring tools produce bias "to some degree," yet 51% of companies still use AI in recruiting. The gap between perceived bias and adoption suggests most hiring teams prioritize speed and cost reduction over bias mitigation.

How many lawsuits are actually happening?

At least three federal cases are active or settled. EEOC v. iTutorGroup (2023, $365K settlement) addressed age discrimination via AI video-interview screening. Mobley v. Workday (2024–2025) conditionally certified a collective action on behalf of all applicants 40+ years old rejected by Workday AI since 2020 — potentially covering hundreds of millions of rejected applications. ACLU v. Aon (EEOC charge filed Dec 2023, ongoing) alleges Aon's AI personality and cognitive assessments screen out neurodivergent and Black applicants. Settlement amounts and timelines are not yet public.

If I apply to a role that uses AI screening, what should I do?

Treat AI screening as a filter, not a gatekeeper. First: personalize your résumé for each role — the Resume.io 2025 study found 62% of hiring managers reject AI-written résumés *without personalization*. Generic prose triggers rejection regardless of AI use. Second: use synonyms for your skills and role titles — AI matchers are sensitive to keyword overlap, and if the job description says "project management" but you wrote "project coordination," a rigid keyword matcher may miss you. Third: include measurable outcomes and context — "40% reduction in onboarding time via process redesign" signals to both AI and human reviewers better than "improved efficiency." Fourth: apply early — Workday data shows application volume grew 4× faster than job openings in 2024, so later applications face steeper ranking. None of these tactics are AI-specific; they all improve your odds against human reviewers too.

Should I disclose that I used AI on my résumé?

No. Resume.io's 2025 survey found 49% of hiring managers say they would reject a résumé they *identified as* AI-generated, but the rejection trigger is generic, impersonal language — not AI use itself. The NBER study showed AI-*assisted* résumés (human-written, AI-polished) had a +8% hiring lift. Disclosing "written with ChatGPT" introduces risk without upside: the screener or hiring manager may not distinguish between human-drafted-AI-polished and AI-generated-wholesale. Keep it simple: use AI to improve grammar, clarity, and relevance, but ensure the voice and examples are yours.

CareerCareer Match Big Five RIASEC

AI Hiring Bias Studies 2024–2025: The Evidence, Legal Cases, and What Job-Seekers Need to Know

|May 4, 2026|11 min read

Take the Career Match test3 min · freeStart

For teams & organizations

Free assessments for HR teams

56 tests for hiring, onboarding, and team development. No per-employee fees, no certification.

Learn more

References

Wilson, K. & Caliskan, A. (2024). Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrieval
University of Washington News (2024). AI tools show biases in ranking job applicants by name
Seyfarth Shaw (2025). Mobley v. Workday — AI service providers could be directly liable for employment discrimination
US EEOC (2024). Employment Discrimination and AI for Workers
US EEOC (2023). EEOC v. iTutorGroup — AI Age Discrimination Settlement
ACLU & EEOC (2023). ACLU v. Aon — AI Assessment Bias Complaint
Resume.io (2025). Resume Rejection Survey 2025
ResumeBuilder (2024). ResumeBuilder AI in Hiring Survey 2024
Workday (2024). Workday Global Workforce Report (H1 2024)
MIT Technology Review (2024). AI Hiring Tools Show Consistent Racial Bias

Ready when you are

Find your ideal career match in 3 minutes.

12 questions, instant result, free forever. No email, no signup — just answer and see your type.

Take the Career Match test Not sure? Browse all 56tests →

Explore Remote Careers

🌍Project Manager→📣Digital Marketing Manager→🔍Recruiter→All careers →

← Back to all articles