news
AI Hiring Tools Develop Their Own Biases, MIT Researchers Find
MIT Technology Review research shows AI hiring tools don't just pick up human biases from training data, they generate their own, too. This means job candidates could face a double-whammy of unfairness before a human ever sees their resume. It's a wake-up call for companies hoping AI will streamline hiring without bias.
MIT Technology Review research shows AI hiring tools don't just pick up human biases from training data, they generate their own, too. This means job candidates could face a double-whammy of unfairness before a human ever sees their resume. It's a wake-up call for companies hoping AI will streamline hiring without bias.
The architecture take
So, you thought AI was neutral? Think again. MIT found these hiring algorithms don't just parrot our prejudices; they invent new ones. It’s a classic case of "Garbage In, Generative Garbage Out," but with a self-modifying twist. For your job search or company culture, this means the 'objective' AI screen could be throwing out perfectly good candidates for reasons no one (not even the AI) fully understands. It's not just about diversity numbers; biased AI affects team cohesion and mental health in the workplace, stressing out the applicants and potentially missing out on the exact nervous-system diversity a team actually needs. What we're watching for? Companies that ditch these tools or invest heavily in transparent, human-in-the-loop oversight to keep the algorithms honest.