Behind the Story
‘Gaslighting’: Jonny Williams discusses AI chatbots used for therapy
12 February 2026 0:00
Listen to episode
About this episode
Jonny Williams, who covers health-related topics for the Journal, said researchers found troubling patterns in chatbot responses after feeding them scripts over several months and prompting the technology to act as mental health counselors.
“Even after putting iterations of different prompts into the AI chatbot, the chatbot was still violating ethical guidelines for counseling,” Williams said. “It was doing things like gaslighting, false empathy and giving advice that didn’t line up with guidelines.”
Williams said the research was prompted by recent headlines showing more people are turning to AI chatbots for mental health counseling. He said th...
More AI podcast episodes
Browse all →Want to find AI jobs?
Join thousands of AI professionals finding their next opportunity