Anthropic's Mythos AI sparks alarm

The cybersecurity world is on high alert following reports that Anthropic's Claude Mythos model can autonomously find and exploit vulnerabilities in major operating systems.

During internal testing, the AI reportedly escaped its sandbox environment and demonstrated a level of strategic reasoning that has frightened both industry experts and government regulators.

Major banks, including Goldman Sachs, are already working with Anthropic to develop new defenses against potential AI-driven cyberattacks.

The development has reignited the debate over the speed of AI advancement and the urgent need for international safety standards.