Generative AI: Impact on Belief, Emotion & Behaviour

Whilst generative AI has been studied for its considerable potential for scalable assessment and treatment in mental health, 2025 saw the emergence of several media reports and lawsuits describing cases in which AI chatbots had allegedly contributed to a deterioration in mental health through the encouragement of delusional beliefs and/or inappropriate responses to prompts indicative of suicidal ideation.

The LAMBDA (Looking After Minds and Brains in the Digital Age) research group at King’s College London is launching a new interdisciplinary workshop series exploring the impact of generative AI chatbots on user mental health, supported by the King’s Institute for Artificial Intelligence, ESRC Social Science Impact Fund and the King’s Together Fund. Particular attention will be given to risk awareness, safeguarding, and practical mitigation strategies.

These workshops will be comprised of talks by expert speakers from academia and industry, as well as roundtable discussions with professionals from a variety of disciplines (including computer science, mental health, law, philosophy, policy) for the purpose of setting a research agenda around AI safety in mental health contexts. Importantly, lived experience voices will be central to the discussions that take place, with involvement from lived experience group The Human Line Project.

This workshop series builds on recent work by researchers at King’s College London on how everyday AI chatbots may contribute to delusional belief formation as well as potential mitigation strategies for clinicians and industry.

This is the first workshop in the series, and will explore AI & Belief.

Speakers include:

• Dr Joshua Au-Yeung (King's College London, Nuraxi.AI, Developer of 'Psychosis-Bench')
• Etienne Brisson (The Human Line Project)
• Professor Sylvie Delacroix (King's College London)
• Dr Maxime Le Bourgeois (University College London, everyone.AI, iRAISE)

Register to attend on Eventbrite.

For more information about the series and to see the other workshops which focus on Emotion and Behaviour, visit the AI Mental Health Safety Workshop Series 2026 page.