IAPQB

When screening software rejects older applicants

Who answers in law when automated screening turns job applicants away, and what employers and vendors now check

IAPQB Report No. 6 · Practice report · Published 17 November 2026
International AI Practice Qualification Board (IAPQB)

In one minute

  • In 2023, iTutorGroup agreed to pay $365,000 to settle a US Equal Employment Opportunity Commission (EEOC) lawsuit. The EEOC said the company's software automatically rejected women aged 55 or older and men aged 60 or older [1].
  • In Mobley v. Workday, a US federal court ruled in 2024 that an HR software vendor could face an applicant's discrimination claims as the agent of employers using its tools [2]. In 2025 it let older applicants join as a group [3]; the case was still open in late October 2026 [4].
  • California now treats anyone who does hiring tasks for an employer, including through software, as an employer [5, 6]. A state audit found gaps in New York City's enforcement of its bias-audit law [7].
  • Courts and regulators point to five practices: questions to vendors, bias testing before and during use, human review on request, records kept for years, and age data left out of screening [1, 3, 6, 7, 8, 9].

1. A familiar situation

A recruiter opens the hiring system on a Monday. Over the weekend, hundreds of people applied for one job, and the software has already sorted them and rejected many, with no reasons shown. If an older applicant later asks why she was turned away, the answer depends on settings the recruiter did not choose and on records the company may not keep.

2. Why it happens

Screening software applies the same rules to every applicant, at high speed [9]. According to the EEOC, iTutorGroup rejected more than 200 qualified applicants because of their age, through a rule in its application software [1]. New York auditors noted that job applicants may not know when such a tool is used [7].

Responsibility is also split between the employer, which chooses the tool, and the vendor, which builds it. In Mobley, the court looked at what the software was alleged to do: take part in decisions "by recommending some candidates to move forward and rejecting others" [2]. It wrote: "Drawing an artificial distinction between software decisionmakers and human decisionmakers would potentially gut anti-discrimination laws in the modern era" [2].

IAPQB view. In both court cases, software allegedly made the first decision about each applicant, and companies had to answer for it: the employer accused of setting an age rule [1] and, so far, the vendor whose software allegedly screened applicants [2, 3]. Both cases turned on what the software did to applicants of different ages, and on who decided how it worked.

3. Three cases

3.1 United States, 2023: an age rule inside the application software

iTutorGroup, a group of three companies offering English tutoring to students in China, hired tutors in the United States. According to the EEOC's lawsuit, the companies "programmed their tutor application software to automatically reject female applicants aged 55 or older and male applicants aged 60 or older", in breach of the Age Discrimination in Employment Act [1].

Under a settlement announced in September 2023, iTutorGroup agreed to pay $365,000 to applicants rejected because of their age. It had stopped hiring tutors in the United States. If it resumes, it must notify and interview those applicants, and the settlement adds training and an order against asking for birth dates. The EEOC will monitor compliance for at least five years [1].

3.2 California, 2024 and 2025: the software vendor in court

In February 2023 a job applicant sued Workday, an HR software company, in federal court in California, alleging race, age and disability discrimination by its screening software [4].

On 12 July 2024 Judge Rita F. Lin held that he had plausibly alleged that Workday acted as its customers' agent, and so could count as an employer under federal anti-discrimination law [2]. The court found it plausible that "Workday's customers delegated their traditional function of rejecting candidates" [2]. Many software vendors are not agents, she added: a spreadsheet program "is not participating in the determination of which employees to hire" [2]. The court dismissed the claims of intentional discrimination and let the claims of disparate impact, an unequal effect on a protected group, go ahead [2].

On 16 May 2025 the court granted preliminary certification of a collective action, a group case others can join, for "all individuals aged 40 and over" who applied through Workday's platform from 24 September 2020 and "were denied employment recommendations" [3].

In July 2025 the court ruled that the group includes applicants screened with Workday's HiredScore AI features, and ordered Workday to identify the customers that used them [4]. The Civil Rights Litigation Clearinghouse lists the case as ongoing; its docket was last updated on 30 October 2026 [4].

3.3 New York City, 2025: an audit of the bias-audit law

In July 2023 New York City began enforcing Local Law 144, which requires yearly independent bias audits of automated hiring tools, a published summary and notice to candidates [7].

In December 2025 the New York State Comptroller reported on the city's enforcement from July 2023 to June 2025. The Department of Consumer and Worker Protection (DCWP) received two complaints. It reviewed 32 companies and found one issue of non-compliance. State auditors found at least 17 potential issues at the same companies, and said DCWP lacked technical expertise to evaluate the tools. DCWP said it was willing to adopt many recommendations, but it disagreed with many findings [7].

4. How practitioners handle it now

Buyers question the vendor and check what the contract allows. Under California's rules, anyone carrying out an employer's usual functions on its behalf, including "in whole or in part" through an automated-decision system, is its agent, and an agent "is also an employer" [6]. In Mobley, Workday argued that its customer contracts barred it from accessing applicant data; the court noted an exception in those contracts for court orders [3]. The UK Information Commissioner's Office (ICO) has published key questions for organisations buying AI recruitment tools [8].

Bias testing happens before use and continues during use. New York City requires an audit before use and every year after [7]. In California, evidence of anti-bias testing, or its absence, is relevant to a claim and to the defence, including the testing's "quality, efficacy, recency, and scope" and "the response to the results" [6]. The ICO told some tool makers it audited in 2024 to check regularly for discrimination [8, 9], and in 2026 advised employers to consider monthly bias reviews [9].

Rejected applicants can reach a person. The UK's Data (Use and Access) Act 2025 made it easier for firms to use automation to make decisions without human involvement [9]. In March 2026 the ICO said organisations must tell candidates how to challenge a decision and request a human review [9].

Records are kept for years. The Mobley group reaches back to applications made from September 2020 [3]. California requires employers to keep employment records, including "automated-decision system data", for four years [5, 6].

Data that reveals age stays out of screening. The iTutorGroup settlement bars asking for birth dates [1]. In Mobley, the court noted that graduation years or work history may serve as proxies for age [3]. The ICO found tools that let recruiters filter out candidates with certain protected characteristics [8].

5. What to take from this

For recruiters and hiring staff. Find out which steps the software decides alone, and how a rejected applicant can ask for a human review [9]. In iTutorGroup, the alleged age limits were software settings, and the company paid to settle [1].

For HR leaders who buy screening tools. Ask the vendor what the tool scores, what bias testing it has had, and what data the company can get back if a court orders it [3, 6, 8, 9]. Keep records at least as long as the strictest rule that applies [6].

For HR software vendors. In Mobley, a vendor whose software allegedly recommends and rejects applicants could be sued as the employers' agent [2]. The ICO audited tool makers directly [8], and the Mobley court ordered Workday to name the customers that used HiredScore features [4].

6. Limits of this report

The cases and rules come from the United States and the United Kingdom; EU law and US federal agency guidance are not reviewed. Both court cases rest on allegations: iTutorGroup settled, and Mobley was still open on the latest docket checked [1, 4]. Mobley facts after May 2025 come from the Clearinghouse summary and docket [4]. The California rules took effect on 1 October 2025 [5]; the text quoted is the Council's final text of 17 March 2025 [6]. This report does not test any screening tool.

References

[1] U.S. Equal Employment Opportunity Commission (11 September 2023). iTutorGroup to Pay $365,000 to Settle EEOC Discriminatory Hiring Suit. https://www.eeoc.gov/newsroom/itutorgroup-pay-365000-settle-eeoc-discriminatory-hiring-suit

[2] Mobley v. Workday, Inc., No. 23-cv-00770-RFL (N.D. Cal.), order granting in part and denying in part motion to dismiss, ECF No. 80, 12 July 2024. https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.80.0.pdf

[3] Mobley v. Workday, Inc., No. 23-cv-00770-RFL (N.D. Cal.), order granting preliminary collective certification, ECF No. 128, 16 May 2025. https://www.govinfo.gov/content/pkg/USCOURTS-cand-3_23-cv-00770/pdf/USCOURTS-cand-3_23-cv-00770-1.pdf

[4] Civil Rights Litigation Clearinghouse. Mobley v. Workday, Inc. 3:23-cv-00770 (N.D. Cal.). https://clearinghouse.net/case/44074/ (accessed 16 November 2026; case summary last updated 17 December 2025, docket last updated 30 October 2026)

[5] California Civil Rights Department (30 June 2025). Civil Rights Council Secures Approval for Regulations to Protect Against Employment Discrimination Related to Artificial Intelligence. https://calcivilrights.ca.gov/wp-content/uploads/sites/32/2025/06/2025.06.30-Automated-Decisions-Regs-Release.pdf

[6] California Civil Rights Council (17 March 2025). Final Unmodified Text of Proposed Employment Regulations Regarding Automated-Decision Systems (Attachment B), sections 11008, 11009(f) and 11013(c). https://calcivilrights.ca.gov/wp-content/uploads/sites/32/2025/03/Attachment-B-Final-Unmodified-Text-of-Proposed-Employment-Regulations-Regarding-Automated-Decision-Systems.pdf

[7] New York State Office of the State Comptroller (2 December 2025). Enforcement of Local Law 144 – Automated Employment Decision Tools. Report 2024-N-6. https://www.osc.ny.gov/state-agencies/audits/2025/12/02/enforcement-local-law-144-automated-employment-decision-tools (full report: https://www.osc.ny.gov/files/state-agencies/audits/pdf/sga-2026-24n6.pdf)

[8] Information Commissioner's Office (6 November 2024). ICO intervention into AI recruitment tools leads to better data protection for job seekers. https://ico.org.uk/about-the-ico/media-centre/news-and-blogs/2024/11/ico-intervention-into-ai-recruitment-tools-leads-to-better-data-protection-for-job-seekers/

[9] Information Commissioner's Office (31 March 2026). Automated decisions can streamline the hiring process – with the right safeguards in place. https://ico.org.uk/about-the-ico/media-centre/news-and-blogs/2026/03/automated-decisions-can-streamline-the-hiring-process-with-the-right-safeguards-in-place/

How to cite

IAPQB (2026). When screening software rejects older applicants: who answers in law when automated screening turns job applicants away, and what employers and vendors now check. IAPQB Report No. 6. International AI Practice Qualification Board. https://iapqb.org/research/006-older-applicants-screening.html

About this report

Prepared by IAPQB. Each fact comes from the sources listed, and each source was opened and checked on 16 November 2026; anything that could not be checked was left out. AI tools were used to search and to draft, and IAPQB checked every claim against its source and is responsible for the text. IAPQB runs an AI work qualification; this report does not promote it. Corrections: contact@iapqb.org or an issue at github.com/iapqb/iapqb.github.io. Licence: CC BY 4.0.

Archived with the series on Zenodo: doi.org/10.5281/zenodo.24248199 · Markdown sources: github.com/iapqb/research