AI hiring tools carry real bias risks, but the answer is better oversight and smarter design. Research shows these systems have complex biases that do not map onto standard categories cleanly, and can be constantly worked on and improved. The fix is rigorous impact assessments, testing and human review, because well-built AI can still outperform the flawed human judgment it replaces.
Workday's AI rejected one man over 100 times, and a federal court just said that's worth a nationwide lawsuit. When even top AI developers admit they don't know why their systems make the choices they do, trusting those systems with hiring decisions is reckless. Employers using these tools are sitting on a legal time bomb, and no efficiency gain is worth discriminating against protected workers.
© 2026 Improve the News Foundation.
All rights reserved.
Version 7.7.2