Anthropic tries to keep pace with OpenAI, faces off with David Sacks

Artificial intelligence startup Anthropic is doing all it can to keep pace with larger rival OpenAI, which is spending money at a historic pace with backing from Microsoft and Nvidia. Of late, Anthropic has been facing an equally daunting antagonist: the U.S. government.

David Sacks, the venture capitalist serving as President Donald Trump’s AI and crypto czar, has been publicly criticizing Anthropic for what he’s called a campaign by the company to support “the Left’s vision of AI regulation.”

After Anthropic co-founder Jack Clark, AI startup’s head of policy, wrote an essay this week titled “Technological Optimism and Appropriate Fear,” Sacks lashed out against the company on X.

“Anthropic is running a sophisticated regulatory capture strategy based on fear-mongering,” Sacks wrote on Tuesday.

OpenAI, meanwhile, has established itself as a partner to the White House since the very beginning of the second Trump administration. On Jan. 21, the day after the inauguration, Trump announced a joint venture called Stargate with OpenAI, Oracle and Softbank to invest billions of dollars in U.S. AI infrastructure.

Sacks’ criticism of Anthropic hits on the company’s very foundation and its original reason for being. Siblings Dario and Daniela Amodei left OpenAI in late 2020 and started Anthropic with a mission to build safer AI. OpenAI had started as a nonprofit lab in 2015, but was rapidly moving towards commercialization, with hefty funding from Microsoft.

Now they’re the two most highly valued private AI companies in the country, with OpenAI commanding a $500 billion valuation and Anthropic capturing a valuation of $183 billion. OpenAI leads the consumer AI market with its ChatGPT and Sora apps, while Anthropic’s Claude models are particularly popular in the enterprise.

When it comes to regulation, the companies have very different views. OpenAI has lobbied for fewer guardrails, while Anthropic has opposed part of the Trump administration’s effort to limit protections.

Anthropic has repeatedly pushed back against efforts by the federal government to preempt state-level regulation of AI, most notably a Trump-backed provision that would have blocked such rules for 10 years.

That proposal, part of the draft “Big Beautiful Bill,” was ultimately abandoned. Anthropic later endorsed California’s SB 53, which would require transparency and safety disclosures from AI companies, effectively going in the opposite direction from the administration’s approach.

“SB 53’s transparency requirements will have an important impact on frontier AI safety,” Anthropic wrote in a blog post on Sept. 8. “Without it, labs with increasingly powerful models could face growing incentives to dial back their own safety and disclosure programs in order to compete.” 

Anthropic didn’t provide a comment for this story. Sacks didn’t respond to a request for comment.

For Sacks, the priority in AI is to innovate as fast as possible to make sure the U.S. doesn’t lose to China.

“The U.S. is currently in an AI race, and our chief global competition is China,” Sacks said in an onstage interview at Salesforce’s Dreamforce conference in San Francisco this week. “They’re the only other country that has the talent, the resources, and the technology expertise to basically beat us in AI.”

But Sacks has adamantly denied that he’s trying to take down Anthropic in the process of lifting up U.S. AI.

In a post on X on Thursday, Sacks contested a Bloomberg story that linked his comments to growing federal scrutiny of Anthropic.

“Nothing could be further from the truth,” he wrote. “Just a couple of months ago, the White House approved Anthropic’s Claude app to be offered to all branches of government through the GSA App Store.”

Rather, Sacks claimed that Anthropic has cast itself as a political underdog, positioning its leadership as principled defenders of public safety while pursuing a public campaign that frames any pushback as partisan targeting.

“It has been Anthropic’s government affairs and media strategy to position itself consistently as a foe of the Trump administration,” Sacks said. “But don’t whine to the media that you’re being ‘targeted’ when all we’ve done is articulate a policy disagreement.”

Sacks pointed to several examples of what he sees as adversarial actions. He referenced Dario Amodei’s comparison of Trump to a “feudal warlord” during the 2024 election. Amodei publicly supported Kamala Harris’ campaign for president.

Sacks also referenced op-eds the company ran opposing key parts of the Trump administration’s AI policy agenda, including its proposed moratorium on state-level regulation and elements of its Middle East and chip export strategy. Anthropic also hired senior Biden-era officials to lead its government relations team, Sacks noted.

The AI czar took particular umbrage to Clark’s essay and his warnings about the potentially transformative and destabilizing power of AI.

“My own experience is that as these AI systems get smarter and smarter, they develop more and more complicated goals. When these goals aren’t absolutely aligned with both our preferences and the right context, the AI systems will behave strangely,” Clark wrote. “Another reason for my fear is I can see a path to these systems starting to design their successors, albeit in a very early form.”

Sacks said such “fear-mongering” is holding back innovation.

“It is principally responsible for the state regulatory frenzy that is damaging the startup ecosystem,” Sacks wrote on X.

Anthropic has also stayed away from actions that many other tech companies have taken explicitly to appease Trump.

Leaders from Meta, OpenAI, and Nvidia have courted Trump and his allies, attending White House dinners, committing tens of billions of dollars to U.S. infrastructure projects, and softening their public postures. Amodei wasn’t invited to a recent White House dinner involving numerous industry leaders, the company confirmed to The Information.

Still, Anthropic continues to hold major federal contracts, including a $200 million deal with the Department of Defense and access to federal agencies through the General Services Administration. It also recently formed a national security advisory council to align its work with U.S. interests, and began offering a version of its Claude model to government customers for $1 per year.

But Sacks isn’t the only influential Republican tech investor voicing his critique of the company.

Keith Rabois, whose husband works in the Trump administration, waded into the mix this week.

“If Anthropic actually believed their rhetoric about safety, they can always shut down the company,” Rabois wrote on X. “And lobby then.”

 WATCH: Anthropic’s Mike Krieger on new model release