New York Governor Kathy Hochul is projecting a striking case of political schizophrenia: while the state imposes a moratorium on new data centers, her administration is aggressively deploying neural networks within government operations. While the physical infrastructure of AI is on pause due to environmental and resource concerns, the software is already burning through the regulatory jungle. The real story here isn't the technology itself, but the radical compression of time: a legislative audit that Hochul originally estimated would take five years of work by a small army of clerks was completed in just two months using LLM systems.
The Economics of Rapid Repeal
This digital audit targeted every state rule and regulation. The AI identifies archaic anomalies, ranging from a ridiculous $25 fee for hunting with dogs to a ban on pregnant women working after midnight. For business, these aren't just curiosities; they represent a direct lever for reducing operational costs. When the regulatory burden is shed through automated search rather than decades of backroom lobbying and committee cycles, the speed of doing business changes fundamentally. As Kathy Hochul stated on Bloomberg’s Odd Lots podcast, she wants to see a government that “isn’t on your back, but on your side.”
"I think every level of government should be using this… I intend to make dramatic changes using the power of AI."
This shift toward rapid-fire deregulation demonstrates that the primary value of AI for authorities lies not in drafting new laws, but in forensics and the deletion of old ones. This serves as a ready-made template for C-suite executives and compliance officers. If a state can sift through thousands of conflicting statutes in a couple of months, the traditional multi-year cycle for updating corporate policies is officially prehistoric. The focus is shifting from labor-intensive verification to strategic risk management: what price is an organization willing to pay for a potential algorithmic error during a total structural overhaul?
Sovereignty and Programmatic Risks
Behind New York’s efficiency lies a dangerous paradox between digital ambitions and physical reality. The data center moratorium paints a future where authorities want to consume AI productivity without taking responsibility for the infrastructure footprint. Furthermore, the legal integrity of automated edits remains a gray zone. If an AI mistakenly flags a vital safety rule as obsolete and it is silently deleted, the chain of liability between the software developer, the state agency, and the affected business is severed. For any company implementing similar systems, a critical question arises: who bears the legal burden when an algorithmic audit "denazifies" a regulation that was actually protecting the structure from systemic risk?
If the state continues to block the construction of the hardware that powers these models, a reasonable question follows: where will the data and computing power for Hochul’s next round of “dramatic changes” actually reside?