Anthropic launched Claude Mythos Preview, a cybersecurity-focused model, apparently softening its frozen relationship with the Trump administration and Pentagon.
Anthropic released Claude Mythos Preview, a cybersecurity-specialized model aimed at government use cases, amid a bitter public feud with the Trump administration that began in late February. The DoD had labeled Anthropic a 'supply chain risk' after the company refused to support domestic mass surveillance or fully autonomous lethal weapons. Anthropic filed a lawsuit against that designation and obtained a temporary injunction. CEO Dario Amodei attended a White House meeting Friday, suggesting diplomatic progress, with multiple federal departments and agencies reportedly interested in Mythos.
Claude Mythos Preview signals Anthropic is building a hardened, government-grade model tier — separate from its consumer and commercial APIs. If this model reaches classified networks again, it creates a two-track Claude ecosystem: consumer Claude and defense Claude. Developers building for federal contractors or dual-use security tools need to track which tier they'll have access to and what policy guardrails Mythos enforces differently from standard Claude.
If you're building security tooling that could serve federal clients, test Claude's current API on a red-teaming or vulnerability triage prompt today to establish a baseline before Mythos access specs are published.
Go to claude.ai and open a new conversation
Tags
Also today
Signals by role
Also today
Tools mentioned