AI Chatbots Advise Women Lower Salaries Per Cornell Study

Recent studies reveal that AI chatbots commonly used for salary negotiation advice exhibit significant gender and racial biases, often recommending lower pay for women and minorities compared to men with identical qualifications. Research led by Ivan P. Yamshchikov and others found that popular language models like ChatGPT and Claude consistently suggest lower starting salaries for women, with disparities as large as $120,000 in comparable roles. This bias stems from the underlying training data, which reflects historical inequalities and societal prejudices, resulting in softer negotiation tactics recommended for women, potentially discouraging assertiveness and exacerbating existing pay gaps. Additionally, research from Harvard Business School shows women use AI tools at work 25% less than men, partly due to risk aversion influenced by societal penalties, which may further limit their career advancement and widen pay disparities. Experts warn that lower AI usage among women could also impact the future development of AI technologies, as underrepresentation may lead to systems that inadequately address women's needs. These findings underscore the urgent need to address embedded biases in AI tools and promote equitable AI adoption to avoid reinforcing systemic inequalities in the workplace.