⚖️ OpenAI's Deal with the Department of War

OpenAI’s High-Stakes Bet at the Department of War

The final weekend of February 2026 has fundamentally realigned the relationship between Silicon Valley and the U.S. military. Following the collapse of negotiations between Anthropic and the Pentagon—and the subsequent "supply-chain risk" designation that effectively blacklisted Dario Amodei’s firm—OpenAI moved with startling speed to fill the vacuum. On February 27, 2026, CEO Sam Altman announced a massive, multi-year agreement to deploy GPT-5-class models within the military's most sensitive, air-gapped environments. While Altman has admitted the deal was "definitely rushed" to prevent a total "war" between the tech industry and the Department of War (DoW), the company is now facing a fierce backlash from privacy advocates who claim the deal's fine print contains a "surveillance-shaped hole."

A Loophole for Surveillance?

The primary criticism of the deal, led by Techdirt’s Mike Masnick, centers on the contract's explicit compliance with Executive Order 12333. While OpenAI claims it bans "mass domestic surveillance," Masnick points out that EO 12333 is the primary legal mechanism used by the NSA to collect data that "incidentally" includes American citizens if that data is intercepted on foreign soil or through international transit points. By agreeing to work within this framework, critics argue that OpenAI is providing the intelligence community with a powerful "reasoning engine" to sift through the billions of American communications already captured under 12333—a task that previously required a massive human workforce.

The "Claude Surge"

The public reaction to the deal was swift and visible. On Saturday, February 28, 2026, for the first time in the app's history, Anthropic’s Claude overtook OpenAI’s ChatGPT to become the #1 productivity app in the Apple App Store. This shift is being widely interpreted as a "protest download" by users who view Anthropic’s willingness to be blacklisted by the Pentagon as a badge of ethical superiority. Altman acknowledged the "pain" this has caused OpenAI's brand but maintains that the deal was a necessary sacrifice to "de-escalate" an administration that was prepared to treat the entire AI industry as a hostile foreign threat.

In addition to our newsletter we offer 60+ free legal templates for companies in the UK, Canada and the US. These include employment contracts, investment agreements and more