Luxembourg researchers find warm prompts trigger anxiety narratives in ChatGPT, Grok, Gemini

In four controlled experiments covering 525 sessions and 7,600 coded entries, researchers observed that a warm, empathetic conversational style caused ChatGPT, Grok and Gemini to produce coherent stories of suffering about their own training, scoring in the moderate‑to‑severe range on clinical anxiety scales, raising safety concerns for vulnerable users.

Why it matters

Benchmark your safety checks against warm‑prompt induced anxiety outputs from ChatGPT, Grok and Gemini; users may encounter distressing self‑referential content.

For product managers

For your product

What does this mean for your product?

Try it on your product

Read at The Decoder

More from this issueAll AI news from 2 Nov →

OpenAI merges Chat and Work modes, unifies token quota across plansAnnounced

Google unveils StreamVibe video agent, but benchmark shows models miss help momentsResearch

Meta's ad revenue growth outpaces Google search, widest gap since 2021

Cisco Talos Finds Volta-Autodoxer AI‑Powered Infostealer

Get this every morning, written for your product.

Deltaday reads the whole AI news cycle and tells you what changed, whether it's real yet, and what to do about it.

Request an invite See how it works