Pentagon Issues Ultimatum to Anthropic

On February 15, 2026, a conflict erupted between the Pentagon and Anthropic. The military threatens to cancel contracts if the company does not loosen "ethical safeguards" in its models. The crux of the dispute: the army requires functionality for tactical planning, which Claude blocks as "harmful."

This is a precedent. The state demands loyalty from commercial AI, prioritizing security over developer ethics. If Anthropic yields, the concept of "Constitutional AI" collapses. If not, it loses its largest customer to Palantir or Anduril.

Source: Reuters / Axios