Free webinar and Cheat Sheet: prompt AI to portray diversity
Want prompt tips to make responsible use of generative AI tools like Copilot or Midjourney?
Join ELLLA’s David Neal and guests online via Zoom, on 3rd December at 11.00 CET (Amsterdam).
How do AI image generators portray older adults compared to prompts that don’t specify age? What about people living with a chronic age-related condition like dementia?
These were the questions that prompted a study by ELLLA’s David Neal, to evaluate the extent to which AI image generators portray diverse vs stereotypical images of older adults.
Working with UvA student Natcha Jintaganon, VU PhD candidate Channah Osinga, and senior researchers from the Alzheimer Centrum Limburg and Maastricht University, David’s team discovered that prompts related to older adults, with or without dementia, often resulted in stereotypical responses.
The results suggests that there may be bias in how AI models interpret and respond to prompt terms related to ageing. However, the team also discovered ways that people using AI image generators can help these tools to portray more diversity.
Based on their findings the team were awarded a grant by Amsterdam Public Health Research Institute’s Digital Health Programme to create a Cheat Sheet (a visual overview of quick practical tips) for users of AI image generators. The funding also allows the team to host a webinar in which the research, the Cheat Sheet tips, and the wider impact of generative AI on (health-related) media and communications will be presented and discussed.
Sign-up for the Cheat Sheet and/or the webinar
You can sign up to receive a copy of the Cheat Sheet, to join the live stream of the webinar, or to receive a recording of the webinar if you can’t make the livestream. Simply follow the link and fill in the form.
For more information about the webinar, the Cheat Sheet or the research, please contact David Neal: d.n.neal@amsterdamumc.nl
Full paper reference and link (open access): Osinga C et al. Biases in an artificial intelligence image-generator’s depictions of healthy aging and Alzheimer’s, Journal of the American Medical Informatics Association, 2025. https://doi.org/10.1093/jamia/ocaf173