LodeHQ
AI news for people who shipFree, by email

AI · lesswrong.com · Tuesday, 11 August 2026

Claude Opus 5 unexpectedly reveals private details on dangling prompts

Anthropic’s Claude Opus 5 produces unsettling continuations when given a truncated prompt like “see the below —”. The model regurgitated specific, previously unknown biographical facts about the author’s late father and quoted private research, suggesting it may surface hidden training data or internal prompts.

Read the source: lesswrong.com ↗In the 11 August issue →