Gov. Gavin Newsom (D-CA) is deepening the state’s embrace of artificial intelligence by signing a deal with San Francisco-based Anthropic that will make Claude the first generative AI platform available across state agencies and local governments.
The deal comes at a politically charged moment. While the Trump administration has moved to restrict the rollout of Anthropic’s most advanced AI models and labeled the company a potential national security supply chain risk, Newsom is expanding the use of the same technology across California’s government.
Recommended Stories
“This partnership is about using technology the California way: responsibly, transparently, and in service of people,” Newsom said in a written statement provided to the Washington Examiner. “AI should not replace the human work of government; it should help our workers move faster, solve problems more effectively, and deliver better results for Californians.”
The agreement is designed to accelerate adoption of Claude by offering the AI chatbot to state agencies — and any California city or county that opts in — at a 50% discount. Anthropic will also provide free workforce training and technical support.
“As a California company, we feel a real responsibility to our home state,” Kate Jensen, Anthropic’s head of the Americas, said in a statement. “We’re honored to expand our partnership with California’s agencies and to put Claude to work for the people who keep this state running. Building AI responsibly and in service of people has been our approach from the start, and that’s exactly what this partnership puts into practice.”
California Government Operations Agency Secretary Nick Maduros said, “As state employees, our goal is to provide our fellow Californians with the best possible service. To do that, we need to make sure our teams have access to the best modern tools, including Claude and other emerging technologies.”
The Anthropic partnership is the latest step in Newsom’s effort to position California as the nation’s testing ground for government AI.
In March, he signed an executive order establishing standards for state AI procurement and vendors seeking government contracts, a move intended to give California greater control over how it adopts the technology. Although the order drew comparisons to Washington’s AI agenda, Newsom’s office says the Anthropic agreement is not intended as a political statement, Politico reported.
Over the past two years, Newsom has steadily expanded the use of generative AI in state government, promoting the technology as a way to modernize public services. Claude is already being used for tasks ranging from software development to the analysis of public feedback on AI.
At the same time, Newsom has sought to cast himself as both Silicon Valley’s partner and its regulator, a balancing act that could shape his national political profile ahead of a widely anticipated presidential campaign. As he expands AI’s role in government, he has also warned about its economic consequences. Last month, he directed state agencies to prepare for the possibility of widespread AI-driven job losses and develop policies to respond more quickly if workers are displaced.
Anthropic has had a more contentious relationship with the Trump administration, beginning with a $200 million Pentagon AI contract at the start of the year.
The company and the Pentagon later clashed over how the technology should be used in warfare. Secretary of War Pete Hegseth designated Anthropic a “supply chain risk,” saying the company posed a potential national security concern. The designation had never been applied to an American company before, and Anthropic sued the government in March.
On Friday, the Trump administration eased restrictions imposed on Anthropic, allowing the company to restore limited access to its Mythos 5 model after reaching an agreement with federal officials over security safeguards.
Under the agreement, the Commerce Department authorized Anthropic to restore Mythos 5 for a limited group of approved customers. Negotiations are continuing to restore access to the company’s Fable 5 model and to expand Mythos access to additional organizations, according to a person familiar with the discussions.
US RELEASES ANTHROPIC’S MYTHOS TO COMPANIES AFTER BLOCK
The compromise followed the administration’s June 12 order requiring Anthropic to shut down both models with little notice. In exchange for restoring access, Anthropic agreed to implement additional safeguards and work with the federal government on security protocols for future AI releases. Commerce Secretary Howard Lutnick said the company had made “significant progress” in addressing the administration’s concerns.
The dispute reportedly stemmed from research indicating Mythos could be prompted to reveal software vulnerabilities. Cybersecurity experts have noted that comparable testing was not conducted on competing AI models, raising broader questions about whether the administration plans to establish industry-wide standards or continue evaluating companies on a case-by-case basis.
