Making sense of Anthropic’s fight with the Pentagon—and OpenAI’s opportunity
“Turn on to politics, or politics will turn on you.”
That’s a famous line from perennial presidential candidate Ralph Nader from more than 20 years ago, and it echoes back to what Greek general Pericles said around 420 BCE. Politics will find and change you, whether you want it to or not. Which brings us to Anthropic.
Anthropic—valued at $380 billion, with a deluge of Silicon Valley’s most prominent investors on its cap table—is in what my colleague Jeremy Kahn calls “the biggest crisis in its five-year existence.” The company’s been in an all-out battle with the Pentagon, a standoff that goes something like this: Anthropic has refused to allow its tech to be used around mass surveillance or lethal autonomous weapons. Secretary of War Pete Hegseth didn’t accept this, saying that technology should be used for “any lawful purpose.” And Anthropic didn’t yield.
Cue the cascade of consequences: The Pentagon terminated its $200 million contract with Anthropic and labeled the LLM giant a “supply chain risk.” This is possibly “serious blow to Anthropic’s business” and has no precedent, as Jeremy points out:
Legal and policy experts said the government’s unprecedented decision presents profound questions about the relationship between the government and business in the U.S. It is the first time the U.S. has ever designated an American company a supply chain risk, and the first time the designation has been used in apparent retaliation for a business not agreeing to certain contractual terms. Anthropic said in a statement Friday that it would take legal action to try to overturn the Pentagon’s designation.
Effectively, Anthropic’s door into the trillion-dollar defense industrial complex seems to have raucously slammed closed. And, as Fortune’s Sharon Goldman scooped last week, OpenAI has stepped in to fill the void. Sam Altman seemingly got the deal done rapidly, and Sharon found out what he told employees:
Altman told employees at the all-hands that the government is willing to let OpenAI build its own “safety stack”—that is, a layered system of technical, policy, and human controls that sit between a powerful AI model and real-world use—and that if the model refuses to perform a task, then the government would not force OpenAI to make it do so.
Still, even Altman knows what this looks like, admitting in an “Ask Me Anything” session on X that the deal “was definitely rushed, and the optics don’t look good.” Meanwhile, the chaos continues, as Anthropic’s Claude passed ChatGPT in the Apple App Store over the weekend, and seemingly had an outage this morning. I suspect it’s going to be an even busier week than usual in AI.
See you tomorrow,
Allie Garfinkle
X: @agarfinks
Email: alexandra.garfinkle@fortune.com
Submit a deal for the Term Sheet newsletter here.
Lily Mae Lazarus curated the deals section of today’s newsletter. Subscribe here.
This story was originally featured on Fortune.com