The world’s most popular AI models may be deepening gender bias by consistently portraying women as younger and less experienced than their male counterparts, with a new study raising serious questions about the reliability of AI within the hiring process.

After analysing 1.4 million online images and videos, along with the output of nine large language models, researchers from the universities of Stanford, Oxford, and UC Berkeley found that women are routinely presented as younger than men across thousands of professional and social roles.

By examining content from platforms like Google, Wikipedia, IMDb, Flickr, and YouTube, researchers uncovered a recurring pattern of women appearing as younger and less experienced, even when shown in identical jobs.

Likewise, the study found that mainstream algorithms are amplifying age-related gender bias. For example, when evaluating nearly 40,000 resumes, researchers discovered that ChatGPT assumed women were younger and less qualified than older male applicants, and rated older men more highly than women for the same positions. 

This result appeared whether the researchers provided names or ChatGPT generated its own applicants, which the research team claims is evidence that gender bias is deeply embedded in the system.

The distortion was most stark for high-status, high-earning occupations, like chief executives, surgeons, or astronauts, even though US census data shows no systematic age differences between men and women in the workforce over the past decade.

The study found the same issues when shifting their analysis from images to text. Researchers looked at the relationship between gender and age using billions of words from across the internet, including Reddit, Google News, Wikipedia, and Twitter, and found that words related to youth are much more closely tied to women.

“One concern people might have is that images and videos are kind of unique in that people can wear makeup or apply filters, using image-specific strategies to make themselves look younger. That’s why we also looked at text, and we found exactly the same pattern,” said Professor Solène Delecourt, co-author of the study.

“Our study shows that age-related gender bias is a culture-wide, statistical distortion of reality, pervading online media through images, search engines, videos, text, and generative AI.”

This distortion means that AI models like ChatGPT, which now reportedly counts 800 million active users every week, are depicting a ‘deeply inaccurate picture of the world’, and creating a problematic feedback loop.

The study found that people who viewed occupation-related images online often internalised the biased age cues, which reinforced these stereotypes and fuelled a self-fulfilling cycle that the researchers said could deepen inequality in the labour market.

Recommended reading

• 40% of Tech Pros Expect to Make at Least 3 Career Changes
• Recruitment Leaders Struggling to Find AI Talent, Report Claims
• Is AI Now Essential in Recruitment?

“This is of particular concern given the internet is increasingly how we learn about the social world,” said Professor Douglas Guilbeault, co-author of the study.

“People are spending more time online, and we rely on algorithms that curate information. And so, what if these biased beliefs are spreading and becoming a self-fulfilling prophecy? Our study shows that they are reinforcing stereotypical expectations about how the world should be.”

The findings cast serious doubt on the increasing use of AI in hiring, and reinforce the fears of job applicants who are reluctant to trust machines. 

Recent figures show that 65% of organisations use AI in recruitment, with nearly a fifth (18%) of hiring managers using AI to help make final hiring decisions. However, at least once at the final stages, applicants are more likely to get human eyes on their CV, something far from guaranteed, with 82% of companies using AI to review resumes.