Anthropic Cuts OpenAI's Access To Claude As It Calls Out NVIDIA & Meta's CEOs

In the increasingly competitive AI landscape, companies are growing more protective of their AI models and how they’re used, especially by rival businesses.

That tension has now boiled over, and a new rift has surfaced between two of the industry's biggest players: Anthropic and OpenAI.

According to a recent report, Anthropic has revoked OpenAI’s access to its Claude family of AI models. The move comes after OpenAI allegedly used Claude in combination with internal tools to evaluate performance across areas like coding, writing, and safety, directly comparing it to its own models.


An Anthropic spokesperson said, “OpenAI’s own technical staff were also using our coding tools ahead of the launch of GPT-5,” calling it “a direct violation of our terms of service.” Anthropic’s commercial terms prohibit the use of Claude in ways that support the development of competing services.

OpenAI will still be allowed limited access for “benchmarking and safety evaluations,” with the ban underlining Anthropic’s stance on access restrictions based on OpenAI’s breach of terms.

This is not a new position for Anthropic, as its Chief Science Officer, Jared Kaplan, had earlier justified cutting off access to Windsurf (a rumored OpenAI acquisition target back then, finally being acquired by Google) by stating, “I think it would be odd for us to be selling Claude to OpenAI.”

Responding to the decision, an OpenAI spokesperson said the company’s usage was “industry standard” and added, “While we respect Anthropic’s decision to cut off our API access, it’s disappointing considering our API remains available to them.”

What started as a battle over access to Claude has now unfolded into a clash of ideologies and top personalities in the AI world.

Along with the move to block OpenAI’s access, Anthropic’s CEO Dario Amodei has publicly challenged NVIDIA’s Jensen Huang, accusing him of lying and misrepresenting his views on AI safety.


In a candid interview on the Big Technology podcast, Amodei passionately rejected claims that he wants to control the AI industry or slow down its progress, saying, “I get very angry when people call me a doomer … When someone’s like, ‘This guy’s a doomer. He wants to slow things down.’”

Amodei was responding to growing criticism, including suggestions from Huang that Anthropic is using safety as a smokescreen for dominance. He firmly pushed back, calling that notion “the most outrageous lie I’ve ever heard.”

Throughout the interview, Amodei positioned himself not as an alarmist, but as someone who deeply believes in AI’s potential, provided it’s built and governed with care. “We can have such a good world if we get everything right,” he said, urging the industry to aim for a “race to the top,” not one driven by speed alone.

The conversation didn’t stop there. Amodei soon started a war of words with another industry titan, Meta’s Mark Zuckerberg.

On the same podcast, Amodei took a jab at Meta’s high-stakes talent hunt, calling out the company’s sky-high pay offers for top AI engineers. “If Mark Zuckerberg throws a dart at a dartboard and it hits your name, that doesn't mean you should be paid ten times more than the guy next to you who's just as skilled,” he said.

Amodei suggested that despite Meta’s billion-dollar recruitment push, Anthropic hasn’t seen much impact. “Some [employees] wouldn’t even talk to Meta,” he said, pointing to mission and culture as stronger magnets than money.


While Meta has reportedly offered up to $200 million for a single hire, Amodei remains firm: “We will not compromise our compensation principles or fairness if someone gets a big offer.”

Anthropic has made waves in the last few days for cutting a rival’s access to its AI model, and its CEO criticizing AI leaders. So, is this just the AI competition getting too personal?

Will innovation and ethical values still win over big money in the AI race?

Let us know what you think in the comments section below!