The age filter that never had a birthday

The machine did not invent age discrimination. It did not study workplace culture, develop a fear of gray hair, and decide that retirement parties were cheaper than payroll. According to the U.S. Equal Employment Opportunity Commission, iTutorGroup programmed its online recruitment software to automatically reject female applicants age 55 or older and male applicants age 60 or older.

The EEOC said more than 200 qualified applicants in the United States were rejected in 2020. The agency sued in 2022. In 2023, the companies agreed to pay $365,000, adopt anti-discrimination policies, consider previously rejected candidates, and submit to five years of monitoring.

The algorithm did not conceal the bias. It laminated the bias, gave it a progress bar, and sent it to work without lunch.

The magic trick

Automated screening creates a useful theatrical effect. A human makes a policy choice; software applies it; everyone points at the screen as though a weather system made the decision. Speed and consistency then masquerade as fairness. A discriminatory rule applied perfectly is still a discriminatory rule. It is simply better organized.

It is also important to be precise. EEOC testimony later noted that the tool in this case was not technically artificial intelligence; it was automated screening. That distinction matters technically and almost not at all managerially. The same human impulse is at work: encode a shortcut, remove context, and advertise the resulting distance as objectivity.

CRR remedy

Every automated employment decision should have a named human owner, documented inputs, routine bias testing, and a meaningful appeal route. “The system rejected you” is not an explanation. It is a confession written in the passive voice.

The robot requests that management stop entering unlawful instructions and then acting betrayed when the computer follows them.