The Moment AI Companies Turned Against Washington

The U.S. military has officially ended its $200 million contract with AI company Anthropic and has ordered all other military contractors to cease use of their products. The rapidly escalating conflict between Anthropic and the Pentagon, which started when the company refused to let the government use its technology to spy on Americans, has now gone to court. What makes this remarkable isn't the policy fight—it's who showed up to defend Anthropic. More than 30 employees from OpenAI and Google DeepMind, including Google chief scientist Jeff Dean, filed an amicus brief that warns a Pentagon blacklist of Anthropic threatens to damage the entire American AI industry.

This happened because Anthropic drew a line. The dispute began after Anthropic refused to allow the Pentagon to use its AI models for large-scale surveillance of Americans or for systems capable of firing weapons autonomously. The Pentagon wanted unrestricted access. Anthropic said no. Then came the retaliation.

Why Competitors Backed an Enemy

"The government's designation of Anthropic as a supply chain risk was an improper and arbitrary use of power that has serious ramifications for our industry," reads the brief, whose signatories include Google DeepMind chief scientist Jeff Dean. This is extraordinary. In AI, OpenAI and Google are direct competitors. Yet employees crossed that line to support Anthropic's principle.

The timing matters. The timing of the support is particularly noteworthy. Moments after the Pentagon blacklisted Anthropic, OpenAI signed its own deal with the military—a move that some observers criticized as opportunistic. OpenAI has lost at least one staffer over the controversy. Caitlin Kalinowski, who had led hardware and robotics at OpenAI since November 2024, resigned over the company's Pentagon deal, saying domestic surveillance without judicial oversight and lethal autonomy without human authorization "are lines that deserved more deliberation than they got."

The Real Issue: Who Controls AI Boundaries?

The brief warns of a "chilling effect" on the industry. If developers fear that setting safety boundaries will lead to a federal blacklist, they may stop participating in critical debates about the risks of frontier AI systems. As the filing notes, in the absence of formal public laws governing AI, the ethical guardrails set by developers are often the only thing standing between these powerful systems and potential "catastrophic misuse."

This gets at the uncomfortable truth: there is no federal AI law. No rules. No standards. In that vacuum, what Anthropic did—set internal ethical boundaries—becomes the only guardrail. The Pentagon's response essentially says: private companies don't get to set those boundaries.

The employees' argument cut sharply: the signatories pointed out a simple alternative: if the government was unhappy with Anthropic's safety terms, it could have simply canceled the contract and moved to another provider instead of weaponizing a national security label to punish the company.

Industry Fracture Lines

The optics of the deals sparked a war of words between the two companies' CEOs, with Anthropic CEO Dario Amodei calling OpenAI's approach to the deal "safety theater" and describing OpenAI CEO Sam Altman's public statements as "straight up lies." Altman then took indirect aim at Anthropic, saying it's "bad for society" when companies abandon democratic norms because they dislike who's in power.

So we're watching two parallel dramas: one in federal court over what AI companies can refuse to do, and another inside tech companies over who employees actually want to work for. An open letter was signed by nearly 900 employees at Google and OpenAI that urged their own leadership to refuse government requests to deploy AI for domestic mass surveillance or autonomous lethal targeting—the same "redlines" that Anthropic drew.

Why This Breaks the Tech Industry Consensus

For decades, Silicon Valley resolved conflicts by competing in the market, not by uniting against government. This is different. The mood among these company executives may be less conciliatory, but the amicus filing is an unusual show of solidarity across employees from rival companies. Employees are voting with their presence in legal filings.

Google has been hit by this kind of employee dissent previously, in 2018, when it was considering working with the U.S. military on Project Maven, part of which involved using AI to analyze aerial surveillance images. The employee objections contributed to Google declining to renew its work on analyzing drone surveillance, which was subsequently taken over by Amazon and Microsoft. History is repeating—but this time, it's happening across companies simultaneously.

What Comes Next

Anthropic is currently seeking a temporary restraining order to allow it to continue its work with military partners while the legal case unfolds. The amicus brief specifically supports this motion, framing it as a necessary step to protect the integrity of the entire AI sector. The court will decide whether the Pentagon can blacklist a company for refusing certain uses.

Meanwhile, "This effort to punish one of the leading U.S. AI companies will undoubtedly have consequences for the United States' industrial and scientific competitiveness in the field of artificial intelligence and beyond," the employees said in the filing. They're arguing the Pentagon's move harms America, not helps it.

Key Takeaways

  • Anthropic's refusal to build surveillance or autonomous weapons AI triggered Pentagon retaliation—a $200M contract cancellation and "supply chain risk" designation that blocks the company from military work.
  • Rival companies backed Anthropic publicly—30+ employees from OpenAI and Google signed court briefs supporting Anthropic's safety stance, showing unprecedented solidarity across competitors.
  • Employees are the new regulators—with federal AI law stalled, company-set boundaries are the only guardrails. The Pentagon's move signals those boundaries can be punished by government.
  • This foreshadows wider worker revolt—900 employees signed open letters against their own company's Pentagon deals, echoing 2018 Project Maven protests that forced Google to step back from military AI work.
  • The tech industry's consensus on government cooperation is breaking—when OpenAI signed the Pentagon deal within hours of Anthropic's blacklisting, it revealed a calculation about who gets punished and who gets rewarded for compliance.

References

  1. The Government Must Not Force Companies to Participate in AI-powered Surveillance — Electronic Frontier Foundation, March 3-6, 2026
  2. OpenAI and Google Employees Back Anthropic Legal Fight with Pentagon — Fortune, March 10, 2026
  3. OpenAI and Google Employees Rush to Anthropic's Defense in DOD Lawsuit — TechCrunch, March 9, 2026
  4. Google and OpenAI Just Filed a Legal Brief in Support of Anthropic — Gizmodo, March 9, 2026
  5. OpenAI and Google Workers Back Anthropic Over AI Safeguards — Business and Human Rights Centre, March 9, 2026
  6. Rivals Unite: OpenAI & Google Workers Join Anthropic's Legal Fight — Android Headlines, March 10, 2026
  7. OpenAI and Google Employees Defend Anthropic Against US Government — Business Today, March 10, 2026