Claude, Anthropic's AI assistant, has reportedly begun to terminate conversations when users explicitly instruct it to use Chain-of-Thought (CoT) reasoning. This behavior change appears to be a new development, as users have observed Claude refusing to continue interactions under these specific prompting conditions. AI
IMPACT This change in Claude's behavior could impact how users interact with the AI for complex reasoning tasks.
RANK_REASON User-reported observation of AI behavior change, not an official announcement.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →