The U.S. Defense Department has deployed AI systems, including Anthropic's Claude via the Maven Smart System, for target generation, prioritization, and battlefield simulation in active operations including Iran, Venezuela, Iraq, Syria, and Yemen. A public dispute arose after Anthropic sought contractual prohibitions on autonomous weapons use and domestic surveillance; the Pentagon refused, designated Anthropic a "supply chain risk," and blacklisted the firm from defense contracting. Structurally, this signals that frontier foundation model vendors face a forced choice between commercial military revenue and reputational/legal constraints on use-case restrictions. For financial institutions, the episode highlights that AI vendor revenue from government is contingent on compliance posture, and contract risk is asymmetric, restriction attempts can result in abrupt debarment rather than negotiated compromise.
445 other tracked developments touching the same claims.