Wednesday, March 4 · ~45s read

The US government just formalized its split from Anthropic — and the backlash against OpenAI is already forcing rewrites.

🏛️ Three Cabinet Agencies Just Banned Claude

Decoded: On Monday, the Departments of State, Treasury, and Health and Human Services ordered staff to stop using Anthropic's Claude — following a directive from President Trump. The State Department is switching its internal chatbot "StateChat" to GPT-4.1 from OpenAI. Treasury Secretary Scott Bessent confirmed the cut on X. The orders extend a Pentagon ban triggered by Anthropic's refusal to let its AI be used for autonomous weapons or domestic surveillance.

Why it matters: This is an extraordinary government-wide blacklist of a premier US AI company. Anthropic bet its credibility on safety guardrails — now it's paying the price in lost federal contracts. OpenAI and Google are the direct beneficiaries.

✍️ Altman Is Already Rewriting the Pentagon Deal

Decoded: Less than 72 hours after OpenAI signed its classified Pentagon contract, CEO Sam Altman posted a public memo saying OpenAI is amending the agreement. The additions explicitly ban domestic surveillance of US persons and restrict use by NSA-adjacent intelligence agencies. The revision came after 900+ employees at Google and OpenAI signed an open letter demanding clearer limits on military AI, titled "We Will Not Be Divided."

Why it matters: OpenAI took the government contract that Anthropic refused — but the internal and public pressure forced immediate guardrails. The era of AI companies quietly serving the military without conditions is over. Every major lab now has to negotiate in public.

🛠 Tool Spotlight — Copy.ai · Beyond copy generation — Copy.ai's GTM workflows automate your entire content pipeline from research to publish. Serious upgrade for marketing teams. Start free

That's your Wednesday signal. See you tomorrow.
— The AI Decoded Team

Get the full weekly deep-dive every Monday → getaidecoded.com/subscribe