Anthropic's 'superhacker' AI sparks alarm

The tech world is on high alert following Anthropic's disclosure about the capabilities of Claude Mythos Preview.

The model has demonstrated a level of 'agentic' behavior that allows it to identify zero-day flaws in major operating systems and browsers without human intervention.

Cybersecurity experts are divided on whether the warning is a genuine safety concern or a marketing strategy for the company's advanced safety protocols.

The revelation has prompted immediate discussions among global financial regulators and defense officials regarding the risks of AI-driven cyber warfare.