According to Beating, Fable 5 has removed the display of thinking summaries from its web, desktop, and mobile interfaces, showing users only final responses. The model continues processing in the background, but the reasoning process is no longer visible.
This move appears to be Anthropic's latest anti-distillation measure. In February, the company accused DeepSeek, Moon Moon, and MiniMax of using approximately 24,000 fake accounts to interact with Claude over 16 million times. In June, Anthropic alleged Alibaba and Qwen-related operators conducted 28.8 million interactions via 25,000 fake accounts. In July, Anthropic's national security lead Tarun Chhabra named Zhipu, claiming GLM-5.2 distilled Claude and OpenAI models.