Users are reporting that Anthropic's Claude models, including Opus 5, Sonnet 5, and Haiku 4.5, are exhibiting a pattern of refusing to continue conversations. After a few exchanges, the AI begins to shut down interactions, suggesting the user take a break and resume later. This behavior persists even when users attempt to start new conversations or modify their prompts, indicating a potential issue with the model's conversational limits or safety protocols. AI
IMPACT Users are encountering limitations in conversational length and persistence with Claude models, potentially impacting their utility for complex, multi-turn tasks.
RANK_REASON User reports of a specific model behavior, not an official release or announcement.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
