Governor Gavin Newsom just threw down the gauntlet against the Trump administration’s hands-off approach to artificial intelligence. In a new executive order signed this week, the California governor is essentially telling the federal government: “Thanks, but we’ll make our own calls on AI companies here”. And honestly? It’s a power move that could reshape how tech gets regulated in this country.

Here’s what sparked this: The Trump administration’s Department of Defense slapped San Francisco-based AI company Anthropic with a “supply-chain risk” label last month, effectively locking them out of military contracts. Anthropic’s crime? They had the audacity to include clauses in their contracts preventing the military from using their AI tools for domestic mass surveillance and fully autonomous weapons. You know, reasonable stuff. A judge recently blocked that designation temporarily, but Newsom’s order makes it clear California isn’t waiting around for federal approval to protect its own interests.

Newsom’s executive order does two things at once, which is kind of genius: it puts guardrails on AI while simultaneously pushing state agencies to use it more. The order requires California departments to develop new standards for AI contracts that prevent the technology from generating child sexual abuse material, violating civil liberties, or enabling discrimination and unlawful surveillance. State agencies also have to give employees access to vetted generative AI tools and issue guidance on watermarking AI-generated images and videos.

The broader picture here is wild. California is already home to most of the world’s biggest AI companies and leads the nation in AI regulations. Now, over 20 state departments are developing or testing Poppy, a generative AI assistant for government workers. Courts and city governments are jumping on the AI train too. This could either be really smart or really messy, or both.

The contrast with Washington D.C. couldn’t be starker. While Newsom’s pushing for thoughtful guardrails, Trump has signed executive orders actively discouraging states from regulating AI. His administration’s new AI policy framework basically says “let the market figure it out” and completely ignores issues around bias, discrimination, and civil rights. That’s right, they’re not even pretending to care about whether AI systems discriminate against people.

Newsom’s handling of this is being watched carefully by two very different groups: union leaders who’ve made clear they won’t support his presidential ambitions without stronger worker protections from AI, and tech giants pouring millions into California politics to keep regulations loose. The governor’s walking a tightrope between protecting people and keeping Silicon Valley happy.

This executive order isn’t just about one company or one dispute. It’s California saying it has the right to protect its residents and its tech ecosystem from federal overreach. Whether that holds up legally and actually improves how AI gets used? That’s the question everyone’s watching.