‘AI chatbots have many transformative uses but are known to incorporate bias. No literature has been found on using AI to generate ideas for library outreach to different student gender categories (women, men, and non-binary). We formulated prompts for Claude 2 and GPT-4 to generate library outreach ideas for these gender categories and discovered gender stereotypes and biases. This study has implications for the way academic libraries should utilize AI in their services, resources, and practices, such as carefully examining stereotypes or bias that may be present in AI output and keeping potential biases in mind when encouraging AI use.’
Link: https://www.tandfonline.com/doi/full/10.1080/10691316.2025.2482522?src=exp-la