The CEO of Chanel asked ChatGPT to photograph her team, but was disappointed with the result

Chanel’s second-ever female global CEO, Leena Nair, recently discovered a stark contrast between her commitment to gender diversity and the way AI is portraying leadership in the luxury industry. During a visit to Microsoft’s Seattle headquarters, Nair and her team tried out OpenAI’s ChatGPT, but the results were disappointing: When asked to generate an image of Chanel’s senior leadership team, the AI ​​produced an all-male group dressed in business suits – hardly representative. from Chanel, a brand where 76 percent of employees are women, including the CEO.

Speaking at the Stanford Graduate School of Business, Nair described the AI’s response as surprisingly outdated. “We said, ‘Show us a photo of a senior leadership team from Chanel visiting Microsoft’ – it’s all men in suits,” she said. “It was a 100 percent male team, not even in fashionable clothes,” she added humorously. “So, come on. This is what you have to offer?”

Nair’s trip to Silicon Valley also included visits to Google and other tech companies, all part of Chanel’s strategy to integrate AI into all its operations. Since 2021, Chanel has even rolled out Lipscanner, an AI-powered app that allows users to virtually try on lipstick shades. Despite embracing AI, Nair recognizes that challenges such as bias and accuracy issues are part of the process. With Chanel’s 96 percent female customer base in mind, she sees it as essential to work to remove biases that do not align with the company’s identity.

OpenAI responded to Fortune with a statement acknowledging that addressing AI bias is an ongoing task. “We are continually adjusting our models to reduce bias and limit harmful outcomes,” a spokesperson said. Meanwhile, when Fortune repeated Nair’s question to ChatGPT, the resulting image included both men and women, showing potential improvements but indicating that problems still exist.

AI biases against women in leadership are not new. A 2023 study from UCLA found that ChatGPT and Stanford’s Alpaca model favored traditional gendered language when describing men and women. For example, the terms “expert” and “integrity” appeared frequently in descriptions of male candidates, while words such as “beauty” and “delight” were used more often for female candidates. AI models also show a tendency to assume that traditionally male-dominated roles are filled by men, often automatically using the pronouns “he” and “him” for positions such as doctors or engineers.

Since joining Chanel in 2021 after a three-decade career at Unilever, Nair has increased female representation among Chanel executives from 38 percent to nearly 60 percent, in line with her long-standing commitment to diversity in the workplace. Her appointment marks a break from Chanel’s typically male-dominated leadership history, with only one other woman, Maureen Chiquet, serving as global CEO from 2007 to 2016. Nair also became Chanel’s first Indian CEO, a remarkable achievement for an industry with traditional demographics.

For Nair, tackling AI bias is crucial not only for her own company, but also for the technology industry as a whole. “It is so important that we maintain the ethics and integrity of what we do,” she said. “I talk to my friends in tech all the time, all the CEOs, and say, ‘Come on guys, you need to make sure you integrate humanistic thinking into AI.’”