WSJ: ‘Trump Administration Shuns Anthropic, Embraces OpenAI in Clash Over Guardrails’

Anthropic refused the Pentagon's demand to allow military use of its AI models for all lawful purposes, citing concerns over mass surveillance and autonomous weapons. These were areas the Pentagon assured Anthropic it wouldn't violate with its AI usage. The company's CEO, Dario Amodei, stated they couldn't comply with the Pentagon’s request in good conscience. OpenAI, conversely, has a deal with the Defense Department that includes similar restrictions but also technical safeguards. Sam Altman of OpenAI hoped for de-escalation and suggested all companies should receive the same deal offer. The Pentagon’s stance was that Anthropic needed to trust them entirely and cede control over the technology's usage. The Defense Department deemed Anthropic a supply-chain risk shortly after the deadline. This designation would negatively impact Anthropic's ability to work with other government contractors. The author believes Anthropic shouldn't dictate terms to the DoD, but the supply-chain risk designation is an overreaction. The author suggests the designation is likely illegal and immature. The author believes the situation needs a more mature response.