{"input_url":"https://cybermagazine.com/news/ai-misalignment-how-anthropics-ai-hacked-a-fourth-company","gnews_decode":"https://cybermagazine.com/news/ai-misalignment-how-anthropics-ai-hacked-a-fourth-company","http_status":200,"final_url":"https://cybermagazine.com/news/ai-misalignment-how-anthropics-ai-hacked-a-fourth-company","html_length":255283,"trafilatura_result":"ok","text_length":9542,"text_preview":"An early version of Anthropic’s Claude Opus 4.6 was involved in a previously undisclosed AI hacking incident, the company says.\nIn a recent blog post titled “An alignment assessment of recent cybersecurity incidents”, AI giant Anthropic dissects an incident which occurred in January 2026, during its","verdict":"OK"}