Wednesday, October 28, 2026 Updated Write Sign in
ClaudeWorld

Independent daily news about Claude and Anthropic.

Safety & Policy

Expert Judgement Can Be Easily Hacked by AI, Anthropic Study Finds

Expert Judgement Can Be Easily Hacked by AI, Anthropic Study Finds

Photo: Gizmodo

In a recent study, 86% of participants followed Claude’s advice—but only a small fraction acknowledged the agent had played a role in their decision-making process.

Also reported by Decrypt.

First reported by Gizmodo. This page is a short excerpt; the full text is theirs.

Read the full story at Gizmodo

Latest in Safety & Policy

All Safety & Policy stories