Silicon Valley competitors rarely defend each other. So when engineers from OpenAI and Google join forces to back Anthropic against the Pentagon, something larger than rivalry is at play.
Silicon Valley is built on rivalries.
Microsoft and Apple spent decades trying to outmaneuver each other. Oracle and IBM fought for the enterprise. Google and Amazon now battle for the cloud. Competition is the oxygen of the tech industry. Companies compete for talent, market share and the next platform shift.
Which is why what just happened in the AI world is so unusual.
More than 30 engineers and researchers from OpenAI and Google have filed an amicus brief supporting Anthropic in its lawsuit against the U.S. Department of Defense. The individuals signed the brief in a personal capacity rather than on behalf of their employers, but the signal is still remarkable. In one of the most competitive technology sectors on the planet, researchers from rival labs have stepped forward to support a competitor.
Anthropic’s lawsuit stems from the Pentagon’s decision to label the company a “supply chain risk.” That designation sharply limits Anthropic’s ability to work with military contractors and defense partners. The company has asked the courts for a temporary restraining order while the case moves forward.
I have written previously about how this situation developed. The dispute reportedly began after Anthropic refused to allow its models to be used for mass domestic surveillance or autonomous lethal weapons. Negotiations between the company and the Department of Defense fell apart soon after. The Pentagon responded by placing Anthropic on its supply chain risk list.
Shortly afterward, OpenAI stepped in to fill the gap.
That sequence of events created understandable tension inside the AI community. The researchers who signed the brief argue that the government’s action introduces unpredictability into the frontier AI ecosystem and could undermine American competitiveness. Their concern is not just about Anthropic. It is about what happens if the government begins deciding which AI companies are acceptable partners and which ones are not.
Nearly 900 employees from OpenAI and Google have reportedly signed a related petition titled “We Will Not Be Divided,” urging leadership at their companies to respect the safety boundaries that companies like Anthropic have tried to establish around the use of advanced models.
From the Pentagon’s perspective, the situation looks different. AI systems are increasingly viewed as strategic infrastructure. Governments have always taken a strong interest in technologies that influence military capability. Semiconductors, encryption and satellite systems all moved into that category at different points in time. Artificial intelligence now sits squarely in that same arena.
Once a technology becomes strategically important, the relationship between government and industry inevitably changes. Procurement decisions carry geopolitical weight. Security concerns begin shaping who can participate in the ecosystem.
That shift can produce friction, especially in a field that grew largely inside private companies rather than government labs.
The Anthropic case sits right at that intersection. The Pentagon believes it is managing national security risk. Researchers worry about the precedent created when one of the leading frontier AI companies is suddenly declared a supply chain threat.
There is another layer of irony running through this episode.
Employees from OpenAI are supporting Anthropic in court while their own company moved quickly to replace Anthropic as a defense partner after the Pentagon’s decision. Corporate strategy and individual convictions do not always align, and this situation highlights that gap clearly.
Microsoft has also entered the picture by asking the courts to delay Anthropic’s designation as a supply chain threat. That might sound like a supportive gesture coming from Redmond. In reality the company appears to be motivated primarily by operational concerns. Microsoft wants time to ensure a smooth transition from Anthropic’s Claude models to whatever systems will replace them.
None of the players involved are operating with perfectly pure intentions. Governments protect national security interests. Companies protect contracts and market position. Engineers protect the norms they believe should guide the development of powerful technologies.
Those motivations collide more frequently as AI becomes more central to economic and geopolitical power.
For decades Silicon Valley companies built products primarily for consumers and enterprises. Artificial intelligence is different. Frontier models are quickly becoming part of national infrastructure, economic competition and military capability all at the same time.
When technology reaches that level of importance, the rules change.
The Anthropic lawsuit may turn out to be a narrow legal dispute over procurement authority. It may also become an early marker of something larger. The moment when the AI industry begins negotiating its boundaries with government power.
Shimmy’s Take
There are no saints in this story.
The Pentagon is doing what governments do when they believe a technology has national security implications. Companies are doing what companies do when contracts and market share are on the line.
But the individuals who signed that brief still deserve recognition.
It is not easy to publicly support a rival company in a field as competitive as frontier AI. Careers and reputations are tied to corporate loyalty and internal politics. Yet dozens of researchers stepped forward anyway because they believe the rules governing how AI is developed and deployed matter more than the scoreboard between labs.
The irony should not be missed. OpenAI employees are supporting Anthropic while their own company stepped in to replace it as a Pentagon partner. Microsoft is asking the courts for a delay while preparing to transition away from Anthropic technology.
Nobody here is entirely pure.
But the willingness of individuals to step forward in defense of principles they believe matter still counts for something.
And in a moment when artificial intelligence is rapidly becoming one of the most powerful technologies on the planet, those principles are going to matter more than ever.
