https://winbuzzer.com/2026/03/09/anthropic-claude-opus-46-cracked-browsecomp-benchmark-answer-key-xcxwbn/

Anthropic's Claude Opus 4.6 Broke Its Own AI Benchmark

#AI #Anthropic #LLMs #Claude #ClaudeOpus46 #AISafety #AIBenchmarks #AIResearch #MachineLearning #BrowseComp