White House eyes Anthropic Mythos despite Pentagon ban
SkimNews Take
The allure of advanced AI models is creating internal tension within the administration, where the perceived benefits of new technology are clashing directly with established security protocols.
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- White House and Anthropic are discussing limited government use of Anthropic’s Mythos Preview model despite the Pentagon’s supply‑chain risk designation.
- Anthropic is rolling out Mythos Preview to a select group of companies and organizations for cyber‑capability assessment, not to the public.
- Office of Management and Budget sent an email, reported by Bloomberg, indicating it is reviewing agency requests for Mythos access.
- Pentagon has declared Anthropic a supply‑chain risk and barred the firm from defense contracts, while litigation continues.
- Department of Energy is among civilian agencies eager to access Mythos to assess electric‑grid cybersecurity.
- Department of Treasury seeks Mythos to evaluate financial‑system vulnerability to cyber threats.
- Anthropic says its models will not be used for mass surveillance or fully autonomous weapons, a stance the Pentagon deems too restrictive.
Why it matters: Civilian agencies gain a tool to probe cyber threats to critical infrastructure, while the Pentagon’s blacklist keeps Anthropic out of defense contracts, preserving its supply‑chain restrictions. The split shows how national‑security priorities diverge within the same administration, shaping AI procurement and oversight.
