When Gov. Gavin Newsom stepped to the microphone in Sacramento on June 29, he did not announce a pilot, a task force, or another study. He announced the largest deployment of artificial intelligence by any state government in American history: a partnership with Anthropic that hands every one of California's roughly 300,000 state workers access to the Claude AI assistant at a 50 percent discount, with any city or county in the state free to opt in on the same terms.
The deal makes Claude the first AI productivity tool available to all state agencies through the California Department of Technology's newly built Statewide Information Technology Shared Services (SITeS) portal, a centralized storefront that lets departments buy vetted AI tools quickly and at negotiated prices. Beyond the discount, the agreement bundles in free workforce training and hands-on technical assistance from Anthropic's own engineers.
"This partnership is about using technology the California way: responsibly, transparently, and in service of people," Newsom said in a statement. "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."
What the deal actually covers
The half-price offer applies to Claude, Anthropic's flagship chatbot, for drafting and summarizing documents, analyzing information, and supplementing day-to-day work across agencies. Crucially, the same discount extends to California's cities and counties, meaning the deal's reach could eventually stretch well past the state payroll into thousands of local government offices.
California is not starting from zero. The state has already threaded Claude through several live deployments. It powers Engaged California, what the governor's office calls a "first-in-the-nation deliberative democracy platform" that solicits public input on policy. It was used to build Poppy, an in-house assistant "designed by state workers for state workers" that piloted with more than 2,800 employees across 67 departments and is slated for statewide rollout this month. At the DMV, Claude is being used to cut wait times and improve customer service. At the Department of Health Care Services, the largest Medicaid agency in the country, it supports internal workflows for case handling. And the Department of Technology and the Governor's Office of Emergency Services are pairing Claude Security and Claude Code to scan, triage, and patch the state's own software for cyber vulnerabilities.
For Anthropic, a company headquartered in San Francisco, the arrangement is as much homecoming as commercial win. "As a California company, we feel a real responsibility to our home state. 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," said Kate Jensen, Anthropic's Head of Americas. "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."
State officials framed the SITeS portal as the mechanism that makes such a sprawling rollout feasible. "CDT is partnering with departments across the state to leverage the state's purchasing power to make it easy to procure new tools, fast and for the best price," said California Chief Information Officer and Department of Technology Director Chris Given. Government Operations Agency Secretary Nick Maduros put it more plainly: state teams "need to make sure our teams have access to the best modern tools, including Claude and other emerging technologies."
Why it matters
The scale is the story. No state has put a frontier AI model in front of a workforce this large, and California's decision to route it through a shared procurement portal with pre-negotiated pricing offers a template other states will study closely. If the model works, expect governors elsewhere to copy the structure: a single vendor deal, a central storefront, a training program, and an executive order supplying political cover.
That cover matters, because the partnership lands amid a striking split between Sacramento and Washington. Earlier this year, Anthropic and the U.S. Department of Defense clashed over a contract when the company sought to carve out protections barring the government from using Claude to surveil Americans or operate autonomous weapons without human oversight. Defense Secretary Pete Hegseth refused, the Pentagon signed with OpenAI instead, and the federal government went so far as to label Anthropic a "supply-chain risk," a designation that bars it from working with other Pentagon contractors.
California, in effect, is embracing a company the federal government has blacklisted. Asked whether the Pentagon's risk designation surfaced during negotiations, Given told POLITICO that it "just didn't come up." The deal builds directly on Newsom's March 2026 executive order on trusted AI procurement, which strengthened civil-rights and privacy standards for state AI purchases, and on a follow-on order last month directing agencies to prepare for workforce disruption as adoption accelerates.
There is also a cost subtext. The agreement arrives just as enterprises across the economy are wrestling with runaway AI subscription bills. A 50 percent discount on a per-seat basis, multiplied across 300,000 potential users, is a substantial concession, and one that signals how badly AI companies want government reference customers to anchor their public-sector businesses.
What to watch
The open questions now are about execution and accountability. Watch whether Poppy's July statewide launch goes smoothly and how many of the state's 67-plus departments actually adopt Claude once the novelty fades. Watch how many cities and counties opt in, since local uptake will determine whether this is a 300,000-person deal or a millions-of-users deal. And watch for the first stumble: with Claude touching Medicaid workflows and cybersecurity patching, the state's promised guardrails, transparency reports, and equity reviews will face their first real stress test the moment an AI-assisted decision goes wrong. California has bet that it can scale AI faster than anyone while keeping humans in charge. The next year will show whether both halves of that promise hold.
"As a California company, we feel a real responsibility to our home state. 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."— Kate Jensen, Head of Americas, Anthropic