Claude Opus 5.5's Reverse Engineering Capabilities May Have Been Quietly Nerfed

Users report that Claude Opus 5.5 now triggers a safety guardrail mid-response when asked to reverse engineer intellectual property, suggesting its capabilities may have been stealthily nerfed. Anthropic released no new version of Opus 5.5 but announced a string of Claude-related product updates on October 21, matching reports on r/Claude. A minority of users dispute the change, arguing the classifier has always been unreliable.