New Stanford research shows that AI tools used in hiring may be reinforcing harmful assumptions about women, particularly those in mid-to-late career stages.
In a study of 34,500 AI-generated resumes, identical qualifications were rewritten differently depending on gender. Women’s resumes were consistently framed as younger, less experienced, and less senior than men’s. When the AI later evaluated those same resumes, it rated older men highest, repeating the bias it had already introduced.
As AI becomes more common in recruitment and talent decisions, the study highlights a critical risk: bias does not disappear when processes are automated. Rather, it scales. These findings underline the need for stronger oversight, more transparent tools, and leadership that understands how algorithmic assumptions can quietly shape who is seen, who is shortlisted, and who advances.
This article offers clear evidence and an important reminder that building the future of work requires intentional design, not reliance on systems assumed to be neutral.
Read the full story here: https://news.stanford.edu/stories/2025/10/ai-llms-age-bias-older-working-women-research
Image sourced from the series “Emily in Paris”.
