New York puts AI to work auditing every rule on its books
Governor Kathy Hochul says AI is now combing every New York rule for dead weight, from a $25 dog-hunting fee up. India has been running the same fight by hand.
The News
New York Governor Kathy Hochul has confirmed that her administration is using artificial intelligence to sift through the state's entire body of law in search of rules that have outlived their purpose. Speaking on Bloomberg's Odd Lots podcast, Hochul said her team has started deploying "AI to analyze every single rule, regulation, [and] policy" to flag legislation that no longer makes sense.
The examples she offered were pointedly mundane. One is a $25 fee still on the books for taking a dog hunting. Another is an obsolete stipulation affecting pregnant residents. Neither is a headline-grabbing statute, and that is precisely the point: the volume of small, forgotten rules is exactly the kind of haystack a machine can search faster than a room full of officials.
The timing is striking. Hochul made the remarks shortly after signing a moratorium on new AI data centres in the state, a move that reads as a brake on the industry's physical footprint. Using the same technology to streamline her own government suggests a governor drawing a line between AI as sprawling infrastructure and AI as an internal productivity tool.
Why It Matters
Governments have promised to cut red tape for decades, usually through commissions, consultations and slow manual reviews that stall under their own weight. Pointing a language model at the statute book reframes deregulation as a search problem rather than a political marathon. If it works, the cost of finding a dead rule falls close to zero, and the only remaining friction is the will to repeal it.
The caveat is that spotting an outdated rule is the easy half. The last time public bodies rushed to automate judgement, from benefits screening to fraud detection, the errors landed hardest on people least able to contest them. A model that confidently flags a live protection as redundant is more dangerous than a slow reviewer who misses a few. New York's experiment will be judged less on how many rules it finds than on how carefully humans vet the shortlist before anything is struck off.
Indian Angle
India has been fighting this exact battle, only mostly by hand. The Jan Vishwas (Amendment of Provisions) Act, 2023 decriminalised more than 180 provisions across 42 central laws, and the government has repeatedly pitched compliance reduction as a core plank of its ease-of-doing-business drive. Hochul's approach hints at how the next phase could run: let a model triage the thousands of overlapping rules that human committees never reach.
The infrastructure to try this already exists. The IndiaAI Mission, backed by more than Rs 10,000 crore, is meant to seed exactly these public-sector use cases, and MeitY has been vocal about government adoption. A regulatory audit is arguably safer ground than citizen-facing AI, because a flagged rule still passes through Parliament or a state assembly before anything changes.
For Indian regulators, the read-across is sharper still. RBI and SEBI both sit on dense, layered circular archives where obsolete instructions rarely get formally withdrawn. A supervised model that maps contradictions and dead clauses across those archives could do more for compliance costs than another glossy ease-of-doing-business ranking, provided a human signs off on every deletion.
FAQ
What exactly is New York doing?
Governor Kathy Hochul says her team is using AI to review every state rule, regulation and policy to identify outdated legislation. She cited examples such as a $25 fee for taking a dog hunting. The stated goal is to find redundant rules faster than manual review allows, though repeal still requires the normal legislative process.
Did she not just restrict AI in New York?
Yes. Hochul recently signed a moratorium on new AI data centres in the state. The two positions are not contradictory: one limits the industry's physical expansion, while the other uses AI as an internal tool to streamline government. It signals a distinction between AI as heavy infrastructure and AI as software.
Has India tried anything similar?
India's deregulation push has so far been largely manual. The Jan Vishwas (Amendment of Provisions) Act, 2023 decriminalised over 180 provisions across 42 laws. The IndiaAI Mission and MeitY's push for public-sector adoption could enable a more automated audit, but no equivalent statute-wide AI review has been formally announced.
Where can I read the original report?
The Verge reported Hochul's comments, drawn from her interview on Bloomberg's Odd Lots podcast. The full write-up is linked in the source paragraph below.
This story was reported by The Verge. Read the full original coverage at The Verge.