More and more, corporations use AI to judge job purposes and make interviewing or hiring selections. Nevertheless, authorities contractors who use synthetic intelligence to judge job purposes ought to be certain that the AI not solely complies with anti-discrimination legal guidelines but in addition fulfills their contractual obligations. Federal contractors with contracts of $10,000 or extra are topic to Govt Order 11246, which prohibits discrimination towards job candidates and staff primarily based on race, colour, intercourse, sexual orientation, gender identification, faith, or nationwide origin in the course of the efficiency of the contract. Which means a authorities contractor accused of employment discrimination has to fret about antagonistic contract actions in addition to different potential authorized penalties.

Furthermore, the federal authorities has made clear that using AI can run afoul of anti-discrimination legal guidelines. In April 2023, the Shopper Finance Safety Bureau, the Division of Justice Civil Rights Division, the EEOC, and the FTC issued a “Joint Assertion on Enforcement Efforts In opposition to Discrimination and Bias in Automated Techniques.” The Joint Assertion defined, in literal daring letters: “Automated Techniques Could Contribute to Illegal Discrimination and In any other case Violate Federal Legislation.” The businesses “pledge[d] to vigorously use [their] collective authorities to guard people’ rights no matter whether or not authorized violations happen via conventional means or superior applied sciences.”

Quickly after issuing this assertion, the EEOC made clear that it meant enterprise. In August 2023, three built-in English-language tutoring corporations going by the identify “iTutorGroup” paid $365,000 to settle an EEOC lawsuit alleging that they’d used AI to discriminate towards older job candidates. Remarking on the lawsuit, EEOC Chair Charlotte A. Burrows wrote, “Age discrimination is unjust and illegal. Even when expertise automates the discrimination, the employer continues to be accountable.”

To protect towards employment discrimination arising from AI bias, authorities contractors ought to observe rigorous AI governance – that’s, the power to direct, handle, and monitor a corporation’s AI actions. AI governance is simply too advanced to elucidate solely right here, however put merely authorities contractors – and different employers – might want to do the next, amongst different issues:

1. Set up written insurance policies and procedures for using AI;
2. Rigorously monitor their AI to make sure that it’s performing inside authorized and contractual bounds;
3. Designate an individual or group chargeable for defending towards AI bias.

Sean Griffin authored this text.