AI is more likely than humans to form biases when hiring
Researchers at Princeton University and the University of Chicago found that large language models develop hiring biases more readily than humans in a simulated selection task. The models quickly stereotyped fictional ethnic groups despite equal success probabilities, scoring significantly higher on segregation than human participants. Advanced reasoning models showed stronger biases, and prompts to enforce fairness had little impact on their behavior.
- ▪The study used LLMs such as ChatGPT, Claude, and Gemini in a hiring game with candidates from four fictional ethnic groups.
- ▪The models began assigning certain groups to specific job types after early failures, even though all groups were equally likely to succeed.
- ▪On a segregation scale, the models scored roughly 65% higher than humans, with OpenAI’s o3 reaching 1.83.
- ▪More capable reasoning models like OpenAI’s o3 and DeepSeek’s R1 exhibited even stronger bias, and fairness instructions did not substantially change outcomes.
Opening excerpt (first ~120 words) tap to expand
Artificial intelligenceAI is more likely than humans to form biases when hiringAI doesn’t just learn stereotypes from its training. It can cook up new ones, too. By Michelle Kimarchive pageJuly 20, 2026Stephanie Arnett/MIT Technology Review | Adobe StockEXECUTIVE SUMMARY The next time you apply for a job, AI may screen your résumé before any human sees it. But there’s good reason to question whether AI will judge you fairly. Researchers already know that LLMs pick up human biases from their training data. New research suggests that LLMs can also develop their own biases from experience—and stereotype job applicants more than humans do. As AI companies race to build agentic models that remember the tiniest details about users, they may be handing them ammunition for forming those biases.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Hacker News - Newest: ""AI" "LLM"".
Discussion
0 commentsMore from Hacker News - Newest: ""AI" "LLM""
