Overview
- Anthropic CEO Dario Amodei met Friday with White House chief of staff Susie Wiles to discuss the new Claude Mythos model, according to multiple outlets including Axios, CNN, CNBC, and the Associated Press.
- Parts of the intelligence community and CISA are already testing Mythos, and officials are exploring a limited rollout as Bloomberg reported plans to make a version available to major federal agencies and Treasury sought access.
- Anthropic says Mythos can autonomously find and chain software flaws and has uncovered thousands of high‑severity bugs across major operating systems and web browsers, while the UK AI Security Institute called it a step up in controlled tests but not proven against well‑defended systems.
- The talks unfold during a legal fight after the Pentagon labeled Anthropic a supply chain risk over the company’s refusal to allow use for mass domestic surveillance or fully autonomous weapons, with one court blocking a broad ban and an appeals court allowing parts of the blacklist to stand as cases continue.
- Access remains restricted under Project Glasswing to about 40 firms such as Amazon, Apple, Google, Microsoft, CrowdStrike, JPMorgan Chase, and Nvidia to find and patch flaws, and bank leaders and regulators are assessing risks after warnings from industry figures like Barclays’ CEO about potential threats to legacy systems.