Anthropic Limits New AI Due to Fear of Cyberattacks

The AI company Anthropic is proceeding cautiously with the introduction of its new model "Claude Mythos Preview." The background to this is concerns that the technology could be abused by hackers to exploit specific vulnerabilities in software. Instead of a broad rollout, the company is focusing on a controlled release for selected partners. Mythos model identifies critical security vulnerabilities At its core, it is a powerful AI system capable of specifically identifying vulnerabilities in software. According to Anthropic, the model even discovers...