OpenAI tells California to ditch patchwork AI rules

Imagine trying to win the Space Race if California’s aerospace and transistor engineers had to pause and rewrite their blueprints every time they crossed a state border.

That is the exact historical headache OpenAI is warning about. On August 12, OpenAI published an open letter to California Governor Gavin Newsom, urging the state not to choke tech innovation with a chaotic patchwork of local laws.

Right now, roughly 1,000 AI-related bills are churning through state legislatures across the country. OpenAI argues that navigating hundreds of different, conflicting state rules won't make models any safer—it will just tie builders in red tape while foreign competitors race ahead.

A warning against a "CEQA for AI"

OpenAI’s pitch to Newsom is essentially a reciprocity model. If a frontier AI lab has already signed a safety agreement with a federal agency like the new Center for AI Standards and Innovation (CAISI), or signed an equivalent international standard like the EU's AI Code of Practice, California should automatically deem them compliant.

OpenAI noted that it has already committed to working with CAISI to evaluate frontier model risks, and it has already signed the EU code.

Crucially, OpenAI also urged Newsom to carve out clear exemptions for small startups. Large tech firms can afford to staff massive compliance departments, but early-stage teams will simply get crushed under the legal bills.

OpenAI then went for California's most sensitive political nerve: the housing crisis. It warned that mismanaging AI rules could accidentally create a "CEQA for AI innovation," referring to the 1970 California Environmental Quality Act, a well-intentioned law that famously paralyzed homebuilding and sent housing costs skyrocketing across the state.

The geopolitical card

Beyond compliance friction, OpenAI framed uniform rules as a national security issue.

According to the letter, AI developers in the People’s Republic of China will not follow state-level American laws. If US labs are bogged down by mismatched regulations across 50 states, OpenAI argues, rival nations will be the primary beneficiaries.

Aligning California with federal standards, OpenAI claims, is necessary to make sure the world runs on US-led democratic AI rather than autocratic alternatives. The company also reminded Newsom of its recent California Economic Impact Report, pointing out that AI is already pouring billions into the state’s budget.

Why it matters

For developers and early-stage founders, OpenAI's stance is hard to disagree with. Building a software company is difficult enough without having to track 50 different definitions of algorithmic bias or safety disclosure across every state line. Keeping small startups exempt from enterprise-scale red tape is vital to keeping Silicon Valley alive.

At the same time, it’s worth watching the corporate strategy here. Big tech companies almost always prefer federal oversight to state oversight. Federal rules take years to hammer out, and large labs have the lobbying muscle in Washington to help draft them. State lawmakers, on the other hand, are often faster, bolder, and much harder to control.

OpenAI's playbook is undeniably clever: framing state-level safety checks as a threat to national security and invoking California's traumatic housing history gives Newsom a politically palatable reason to tap the brakes. Whether state lawmakers agree to defer their power to federal agencies remains the real battle.