New York Governor Kathy Hochul says her office used AI to review every rule, regulation, and policy in state law in a matter of months — work she estimates would have taken staff five years. Hochul described the effort on Bloomberg's Odd Lots podcast on July 16, 2026, days after signing a moratorium of up to one year on new hyperscale data centers in the state. The pairing is pointed: New York is throttling the physical buildout of AI infrastructure while quietly deploying AI inside the government itself.
Hochul told the podcast her team is using "AI to analyze every single rule, regulation, [and] policy" to flag outdated legislation for repeal. Examples she cited include a $25 fee required to take a dog hunting and a stipulation that pregnant workers need a permit to work after midnight — the kind of statutory sediment that accumulates across decades and rarely gets cleared without a forcing function.
“AI to analyze every single rule, regulation, [and] policy”— Kathy Hochul, Governor of New York
The compression is the story. Hochul said the review "probably would have taken five years at the staff level," and that with AI her team "did it in a couple of months." Whether the estimate holds under scrutiny is a separate question, but the direction of travel matters: a US state executive is publicly attributing a 30x-plus productivity jump on a specific bureaucratic task to language models.
Key facts
- 01Hochul says AI reviewed every New York state rule, regulation, and policy in a couple of months.
- 02The same review would have taken 5 years at the staff level, according to the governor.
- 03New York this week became the first US state to pause new hyperscale data centers, for up to a year.
- 04Hochul cited a $25 fee to take a dog hunting as one of the outdated laws surfaced by the review.
- 05Interview aired on Bloomberg's Odd Lots podcast on July 16, 2026.
This is one of the more concrete examples to date of a US governor operationalizing generative AI on a defined government workflow rather than issuing a task-force report about it. The output is a list of statutes and rules to strike, which then feeds into legislative and administrative action. Hochul did not name the model or vendor powering the review.
The timing sharpens the contrast. Earlier this week, New York became the first US state to pause new hyperscale data centers for up to a year while lawmakers draft rules on utility costs and natural resource impact. Hyperscale campuses are the physical substrate of the frontier AI industry, and a state-level pause — even a temporary one — is a meaningful signal to operators eyeing the Northeast.
Hochul is not framing the two moves as contradictory. In her telling, the moratorium is about protecting residents from rising electricity bills and water draw, while the internal AI review is about making state government leaner. "I want a government that's not on your back but on your side, and using AI has been powerful to do that," she told Bloomberg.
“I want a government that's not on your back but on your side, and using AI has been powerful to do that”— Kathy Hochul, Governor of New York
She went further on the podcast, saying, "I think every level of government should use this… I'm going to make dramatic changes using the power of AI." That is a stronger endorsement of government AI adoption than most governors have offered on the record, and it comes from a state that also just imposed the strictest siting rule in the country on the infrastructure that trains and serves these models.
The unknowns are worth flagging. Hochul's office has not published the methodology, the model used, the prompt structure, or the false-positive rate on the rule review. Regulatory text is the kind of dense, cross-referenced corpus where language models can miss context or hallucinate repeals that would break dependent statutes. Any list of rules to strike will need lawyer review before it moves — the AI is the first pass, not the final word.
The internal review also does not touch the harder question raised by the data center pause: whether states can host the compute the federal government and private industry want built without absorbing the grid and water costs locally. New York is now running a live experiment on both sides of that ledger at once — consuming AI capability while restricting where the capability gets manufactured.
For AI vendors, the takeaway is that state governments are a real market and are moving faster than the federal procurement cycle suggests. A governor citing a specific time savings on a specific workflow is a template other statehouses will copy, and it puts pressure on incumbents in the government-services stack to show comparable numbers. The harder political question — whether a state can be pro-AI-adoption and anti-AI-infrastructure at the same time — is one Hochul is now testing in public.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
