Anthropic Withholds 'Risky' AI Model

Anthropic is currently offering only limited, controlled access to the Mythos model for defensive testing with select partners.

The model reportedly possesses the capability to identify and exploit flaws in major operating systems and web browsers at a scale previously unseen.

This decision has sent shockwaves through the tech industry, triggering a sell-off in cybersecurity software stocks.

The move highlights a growing debate among AI labs regarding the balance between rapid innovation and the potential for catastrophic high-tech risks.