Anthropic rejects US military demands for autonomous weapons

Anthropic declined U.S. government requests in February to lift restrictions prohibiting its AI models from being used in mass domestic surveillance and fully autonomous lethal weapons systems. The dispute, which emerged publicly this week, centers on the company's refusal to modify its acceptable use policy despite alleged pressure from defense officials seeking unrestricted military applications. Anthropic maintains that its safety guidelines are non-negotiable core principles rather than adjustable commercial terms. Government representatives have not publicly confirmed specific demands but have previously advocated for regulatory flexibility in national security contexts. The standoff highlights growing tension between AI companies' published safety commitments and state actors seeking tactical advantages through advanced automation. Industry analysts suggest this conflict could establish precedent for how other foundation model providers negotiate dual-use restrictions with intelligence and defense agencies worldwide.