What Happened
An Uncanny Valley podcast episode from February 26, 2026 covered an ongoing dispute between Anthropic and the Pentagon. According to Wired's reporting, Anthropic turned down a Department of Defense contract, with the conflict rooted in Anthropic's stated safety policies that govern what its models can be used for.
The episode also covers broader topics including agentic versus mimetic AI, and the political backdrop of the current administration's approach to AI governance.
Why It Matters
Anthopic's refusal, if confirmed, is significant for several reasons. Anthropic positions itself as a safety-first AI lab and publishes detailed acceptable use policies. Taking defense contracts that involve lethal applications or classified weapons systems would create an obvious tension with that positioning.
At the same time, other major AI labs have moved in the opposite direction. OpenAI, as of February 2026, signed an agreement with the Department of War (see separate item). The divergence between labs on defense work is becoming one of the clearest differentiators between AI companies.
For enterprise customers in defense-adjacent industries, the question of which AI provider will work with government and military clients is becoming a purchasing criterion, not just an abstract policy question.
Our Take
Anthopic's position is consistent with its public commitments, but it comes with real business consequences. Defense contracts are large and often long-term, and a competitor willing to take that work gains both revenue and deep government relationships.
The more interesting long-term question is whether Anthropic's stance attracts customers who specifically want an AI provider that has drawn clear lines around military applications, or whether it simply cedes that market segment to OpenAI and others. Both outcomes are plausible.