Anthropic's Mythos AI raises alarms

Anthropic has limited the public release of Mythos Preview due to fears that hackers could use the tool to sabotage networks or steal sensitive financial data.

Despite these concerns, the Trump administration is reportedly in talks to give federal agencies access to the powerful tool.

The model represents a significant leap in 'spatial intelligence' and autonomous coding, outperforming previous versions in complex technical tasks.

Critics argue that the rapid deployment of such potent technology without adequate safeguards could lead to catastrophic systemic risks.