Why Does This Matter?
This is widely described in legal and trade press as the EEOC's first-ever lawsuit over discrimination caused by AI hiring software - and it closed with a settlement, not a trial, which shapes what it does and doesn't prove.
Described as the EEOC's first AI hiring discrimination case
Multiple legal-industry outlets that covered the settlement - including Bloomberg Law, which called it the agency's "first-ever AI discrimination in hiring lawsuit," and several employment law firms - describe this as the EEOC's first case built around AI-driven hiring software specifically. Mobley v. Workday, tracked elsewhere on this site, came later and targets a much larger, more complex AI hiring platform; this case came first and involved a comparatively simple piece of software. (source: Bloomberg Law)
A simple rule, not a complex model - and the law still applied
Based on the EEOC's own description, the "AI" at issue here wasn't a sophisticated machine-learning model - it was software that, according to the agency, automatically rejected an applicant based on a simple rule keyed to birth date and sex (reject if female and 55 or older, or male and 60 or older). The case is a reminder that age-old anti-discrimination law - the ADEA has applied to hiring decisions since 1967 - doesn't care whether a human or a script made the discriminatory decision. As EEOC Trial Attorney Daniel Seltzer put it in the agency's own announcement of the lawsuit: "The reach of the laws that the EEOC enforces is long - even fully remote workers providing services to clients abroad may well be employees who are protected from age and other types of discrimination." (source: EEOC press release)
How the discrimination was allegedly caught
According to the EEOC's complaint, as reported by HR Dive, an applicant applied once with her real birth date and was rejected immediately - then reapplied about a day later with identical information except a more recent, younger-appearing birth date, and was offered an interview. If accurate, that side-by-side comparison is a clean, hard-to-explain-away pattern, not just a statistical inference. It's a big part of why this case is used as a teaching example in discussions of how automated hiring bias can be detected. (source: HR Dive)
A settlement, not a ruling - what the consent decree does and doesn't establish
It's important to be precise about what actually happened: iTutorGroup agreed to a consent decree, a settlement that a judge signs off on and can enforce like a court order. Under it, iTutorGroup did not admit that it discriminated against anyone - reporting on the settlement notes the company specifically disputed even the threshold legal question of whether its tutors counted as "employees" under the ADEA at all, arguing they were independent contractors instead. No judge or jury ever ruled on whether the alleged discrimination actually happened. What the consent decree does establish is concrete going forward: a $365,000 payment to the group of rejected applicants, a five-year injunction against age- and sex-based hiring discrimination, a new anti-discrimination policy, mandatory training, and - notably - a flat ban on iTutorGroup asking applicants for their birth date during hiring. (source: EEOC settlement announcement; HR Dive)
To see how this played out step by step, check the timeline.
Sources (all publicly accessible)
- EEOC press release: "EEOC Sues iTutorGroup for Age Discrimination"
- EEOC press release: "iTutorGroup to Pay $365,000 to Settle EEOC Discriminatory Hiring Suit"
- Bloomberg Law — coverage describing this as the EEOC's first AI hiring discrimination lawsuit.
- HR Dive — reporting on the discovery story and iTutorGroup's denial of wrongdoing.