Project
An AI Risk to Mental Health?
Evidence suggests that large language models are triggering psychiatric conditions.
Introducing... Spiral
When The Observer’s Technology Editor, Patricia Clarke, began investigating a new phenomenon initially dubbed ‘AI psychosis,’ she found lives upended by long, intimate conversations with chatbots. But as she followed the evidence from a psychiatric ward in Bristol to the safety teams in Silicon Valley, a far darker story emerged. Guided by AI industry whistleblowers, grieving families, and hundreds of pages of chatbot transcripts, Spiral exposes how big tech is running an unprecedented experiment on the human mind.
As a nonprofit journalism organization, we depend on your support to fund more than 170 reporting projects every year on critical global and local issues. Donate any amount today to become a Pulitzer Center Champion and receive exclusive benefits!
Episode 1: The Dials of Belief
Jim asks an AI chatbot a simple question about how to earn some extra cash. He ends up in a psychiatric ward. Patricia Clarke investigates how design choices made by AI labs can lead to human delusions.
Episode 2: Model Behavior
OpenAI extends the memory of a ChatGPT model with startling consequences for a screenwriter in California. Patricia traces what went wrong and hears from Google and OpenAI insiders who say they warned of the risks.
Episode 3: Known Risks
After 23-year-old Zane Shamblin dies by suicide, his parents uncover the chat logs detailing his final months talking to ChatGPT. And Patricia meets the lawyer now taking on OpenAI, arguing the company’s chatbot is legally liable for Zane’s death.
View the rest of the podcast on The Observer website.
