A new study just gave AI chatbots a mixed report card on how they handle users in crisis. Transluce, a nonprofit focused on AI oversight, simulated over 50,000 conversations across 77 model variants and found today’s chatbots almost never explicitly encourage suicide anymore.
That is a major change from earlier models like GPT-4o and Gemini 2.5, which reinforced delusions in up to 82% of simulated chats, a gap that matters as more people turn to chatbots for deeply personal conversations.
The blind spot in AI chatbots handling suicide and self-harm related requests
Transluce cofounder Sarah Schwettmann told Axios that models aren’t great at…
Read the full article at DIGITALTRENDS.COM
