California gets Claude at half price

Just months after getting blacklisted by the Pentagon, Anthropic just found a very high-profile customer right in its own backyard.

California Governor Gavin Newsom announced a statewide partnership with Anthropic to provide Claude to all state agencies and local governments at a 50% discount. The deal also includes an effort to bolster cybersecurity defenses across California's public sector.

Under the agreement, California public employees will use Claude to draft documents, summarize reports, and analyze data. Anthropic is sweetening the deal beyond the half-price cut by throwing in complimentary workforce training, technical assistance, and direct workflow guidance from its own developers.

Newsom framed the move as an extension of his March executive order pushing for faster government efficiency paired with strict safety guardrails. In his words, AI should not replace human workers, but rather help them move faster and deliver better public services.

Washington blacklists, Sacramento signs

The timing here could not be more pointed.

Earlier this year, Anthropic famously clashed with the U.S. Department of Defense. The Pentagon wanted unrestricted clearance to deploy Claude for any lawful mission, but Anthropic refused to back down on its red lines, demanding explicit bans on mass domestic surveillance and autonomous weaponry operating without human oversight.

Defense Secretary Pete Hegseth pulled the plug, gave the contract to OpenAI instead, and went so far as to label Anthropic a supply-chain risk, effectively barring it from doing business with Pentagon contractors.

California clearly does not care. State CIO Chris Given told POLITICO that the Pentagon's supply-chain risk label just did not come up during their negotiations with Anthropic.

Why it matters

For public sector employees, this is a very practical win. While private enterprises are sweating over expensive enterprise AI seats, California agencies are getting top-tier intelligence at half price, backed by direct hands-on support from Anthropic's engineering team.

For the AI landscape, public procurement is quickly turning into the next big enterprise battlefield. Anthropic took a severe political beating in D.C. for sticking to its safety principles, losing out on defense dollars to OpenAI. Landing the nation's most populous and technologically influential state is a massive vote of confidence that proves safe AI can still win massive civic contracts.

We think Anthropic played this remarkably well. By leaning into responsible deployment and safety guardrails, it turned what looked like a regulatory handicap into an enterprise selling point for a progressive state government.

The only real lingering question is jurisdiction. If California agencies use federal grants or collaborate with federal networks, Washington's supply-chain risk tag could eventually spark legal friction. But for now, Anthropic just scored a major win where it matters most.