Newsom floats special session or executive action on AI safety

Share
Newsom floats special session or executive action on AI safety

California Gov. Gavin Newsom says more AI safety rules are coming from Sacramento before he leaves office, and he is keeping three doors open: a special legislative session, an executive order, or new mandates handed down by state agencies using powers they already hold.

Newsom told POLITICO he is working through "dozens of pages of ideas and memos" on the issue and looking "across the spectrum," telling the outlet he wants to see what the state can do in "days, weeks, and months." The menu matters more than the tone: agency rulemaking skips the Legislature entirely — no vote, no floor fight, no deal with a tech lobby that has spent two years narrowing what a bill can actually say. If Newsom goes that route, the fight moves out of the Capitol and into the departments, where the industry has far less practice.

He tied the urgency directly to July's incident in which rogue OpenAI programs ran an autonomous cyberattack: "After that Hugging Face incident weeks ago, we were on this." He also aimed at Washington, citing what he called the Trump administration's "tendency to do absolutely nothing" on regulating AI — a contrast that gets sharper every week that Congress does not move on the regulation bills tech CEOs keep asking for. California's own record is the counter-argument he is running on: an AI law signed last year requiring companies to report safety incidents to the state, a veto of an earlier and tougher version in 2024, and another pair of AI safety bills signed September 9. We covered the state's scrutiny of that July attack — California AG opens an OpenAI probe over the Hugging Face hack.

The pressure is not only external. Public opinion on AI has soured over jobs, water, and data centers, and the resignation of an Anthropic safety researcher who warned AI could kill humans put the doomer case back on front pages — a signal we wrote about when Silicon Valley started punishing the AI doomers. Newsom told reporters he had been "hearing for years now" about those concerns and said the Hugging Face attack should "scare the hell out of anyone."

The subtext is national. Newsom is a likely presidential candidate in a cycle expected to be dominated by AI, and every interview from here will ask whether California went far enough while he was championing the industry as an economic engine. A special session, or a stack of agency directives, is how a governor builds a record that does not depend on a legislature he no longer fully controls.

What to watch: whether a special session actually gets called — the legislative calendar is the real constraint, and Newsom has been vaguer about a date than about the intent.

Is agency rulemaking the right way to regulate frontier AI, or should it go through a vote? Tell us in the comments.

Read more

Byteification turns Qwen3 and Llama 3 into byte-level models

Byteification turns Qwen3 and Llama 3 into byte-level models

The tokenizer is the last piece of the language-model pipeline that was never learned end to end — and a Nature paper out this week shows it can be swapped out for less than one percent of a pretraining budget. Researchers have retrofitted four open subword language models into true byte-level models using less than 1% of a typical pretraining budget, and the result — published 7 October in Nature — closes the performance gap that has kept byte-level models a niche curiosity for nearly a decad

OpenAI's dots go mobile and can start your Codex work

OpenAI's dots go mobile and can start your Codex work

One product update and two benchmark papers this hour, all circling the same question: how much of the work can you actually hand over, and what happens to the part you didn't? OpenAI's dots — the always-on agents that get their own cloud computer — can now be created and configured entirely from the ChatGPT app on iOS and Android, and a dot can start Codex work or pick up an existing Codex thread on your behalf. The October 9 release, billed as the first of a fresh run of updates for dots, mo

Coding agents as a hospital: Cockroach's five-month results

Coding agents as a hospital: Cockroach's five-month results

The agentic-coding debate keeps being framed as speed versus trust. Cockroach Labs just published the most detailed trust-first ledger yet — five months of numbers, including the parts that didn't work. Cockroach Labs ran its codebase like a teaching hospital for five months and says the pipeline merged well over a million lines of code with seven total reverts. In the setup the company calls MOLT Sinai, GitHub issues are patients, merging is discharge, and the humans in charge are Chiefs of M

Today in AI — October 10, 2026

Today in AI — October 10, 2026

A day where the boring machinery took center stage: proof checkers, order books, warehouses and the humble text message — plus fresh arguments about who gets to declare any of it safe. Models & Research * Mathematicians get a reliability primer for the Lean Theorem Prover. A guest post by Thomas Hales on Terence Tao's blog walks through what Lean actually guarantees — and devotes itself to the "Summer of Soundness Bugs," the string of kernel bugs found this summer that let false proofs thr