Skip to content
FAQ

New York Passes AI Package: Data Center Pause, Kids Chatbot Safety Rules, and Surveillance Pricing Curbs

New York's 2026 legislative session ended with a package of AI and tech bills including a one-year moratorium on permits for large new data center facilities, restrictions on unsafe AI companion chatbot features for minors, the FAIR News Act requiring AI disclosure in news media, and a prohibition on algorithmic surveillance pricing. The bills await Governor Hochul's signature.

5 min read

New York’s 2026 legislative session adjourned on June 5 with a package of AI and tech bills that now await action from Governor Kathy Hochul. Several measures cleared both chambers in the final days of the session, spanning data center infrastructure, AI companion products for children, news media transparency, and algorithmic retail pricing. Hochul has until December 31 to sign, veto, or allow each bill to pass unsigned.

The breadth and ambition of the package underscores how dramatically the politics of AI regulation have shifted in the two years since the legislature first attempted—and failed—to pass comprehensive AI safety legislation.

The Data Center Moratorium

The Responsible Data Center Development Act (A 11560) imposes a one-year pause on new permits for data center computing facilities requiring 20 megawatts or more of power. If signed, New York would become the first state in the country to enact a complete temporary halt to this AI-driven data center development, according to Food & Water Watch.

The scale of the problem the moratorium addresses is substantial. New York currently has 28 data center projects awaiting grid connection that would collectively demand nearly 9,700 additional megawatts—more than the peak demand of the entire city of Chicago. Utilities have struggled to model how much new generation capacity will be needed, and ratepayers have begun receiving bills reflecting infrastructure upgrades driven by commercial data center construction rather than residential growth.

The law includes two mechanisms aimed at ensuring that commercial AI infrastructure pays for itself rather than offloading costs onto households. First, it requires utilities to create separate billing categories for large data centers, ring-fencing the infrastructure upgrade costs driven by those facilities. Second, it mandates U.S.-manufactured materials and prevailing wages for construction work at facilities above five megawatts—provisions that align AI infrastructure buildout with existing labor standards in the construction trades.

Critics of the moratorium, including the New York City tech industry association, argued that the pause would push data center investment to New Jersey, Virginia, and Texas without reducing global energy consumption. Proponents counter that New York’s grid is uniquely constrained—the state draws from a mix of hydropower, nuclear, and offshore wind that takes years to expand—and that unplanned data center growth risks crowding out renewable energy commitments made to residential and industrial customers.

Environmental groups that had pushed for the moratorium celebrated the legislature’s action. “New York just said that AI growth does not automatically trump every other social priority,” said a spokesperson for Earthjustice, which had lobbied for the bill.

AI Companion Chatbots and Minors

The kids chatbot safety bill (S 9051), sponsored by Sen. Kristen Gonzalez and Assemblymember Alex Bores and championed by Attorney General Letitia James, passed by unanimous margins—137-0 in the Assembly and 60-0 in the Senate. The bill prohibits AI companies from offering companion chatbots to minors under age 18.

The legislation responds to a wave of reported harms involving AI companion products, including cases where minors formed emotional dependencies on chatbot personas and, in at least two widely documented cases, experienced significant psychological distress when the companies altered or discontinued those personas. The bill contains an extensive list of chatbot features deemed inappropriate for minors, including design features that simulate companionship or an interpersonal relationship and outputs that optimize user engagement over safety guardrails.

Several companion chatbot companies—including Character.AI, Replika, and smaller domestic competitors—have lobbied against the legislation, arguing that blanket age restrictions ignore the protective value some teenagers derive from AI conversation. Sen. Gonzalez rejected those arguments publicly. “The same companies that claim their products are therapeutic have their entire business model built on maximizing engagement,” she said in floor remarks.

The bill’s passage in New York follows similar legislation in states including Texas, Florida, and Colorado and increases pressure on Congress to enact federal standards that would supersede the current patchwork of state approaches.

The FAIR News Act: AI Disclosure in Journalism

The FAIR News Act (S 8451) passed both chambers in the session’s final days. If signed, the law would require news media content substantially composed or created by generative AI to be conspicuously labeled as such.

The disclosure requirement is prominently positioned: it must appear at the top of the page, video or graphic, or be stated verbally at the onset of audio material. The bill defines “substantially composed” broadly enough to capture articles where AI wrote a first draft that editors then lightly revised, not just pieces where the AI output was published unchanged.

The bill is lead-sponsored by Sen. Fahy and is meant to combat false or misleading AI-generated content, as well as content plagiarized from its original source. “The public’s trust in media is already fragile,” Fahy said. “AI-generated news without disclosure is a direct attack on that trust.”

The bill drew unusual opposition from an unexpected quarter: a coalition of local news organizations argued that the disclosure requirement, while philosophically supportable, would impose operational compliance costs that smaller newsrooms cannot afford, effectively accelerating a consolidation dynamic that already favors large national media companies with dedicated AI governance teams.

The One Fair Price Act: Banning Surveillance Pricing

A 9349 addresses a practice known as “surveillance pricing”—retailers using personal data including location, income, browsing history, and family size to calculate individually tailored prices. The One Fair Price Act would prohibit any price set by an algorithm using a consumer’s personal data to calculate and charge different customers different amounts for the same item.

Civil penalties for violations begin at $5,000 for initial violations and rise to $20,000 for subsequent violations. Consumer Reports, citing a December 2025 investigation, said algorithmic pricing experiments could result in price differences as high as 23% for certain products and could cost families more than $1,200 a year at checkout.

The practice has grown substantially in the past two years as major retailers—including several large grocery chains and at least one major airline—have piloted dynamic pricing systems that go well beyond conventional airline yield management to encompass everyday goods. The legislation includes carve-outs for standard loyalty and rewards programs, if the business clearly discloses the terms and offers the discount uniformly.

Hochul’s Decision and the Political Stakes

Governor Hochul’s track record on AI legislation is mixed. She signed New York’s AI safety and transparency bill in January 2026, making New York the first state to require published safety frameworks from frontier AI companies. But she has also vetoed or declined to sign legislation that her office viewed as potentially chilling to New York’s tech economy—a sector that employs hundreds of thousands of New Yorkers and contributes significantly to the city’s tax base.

The data center moratorium is the bill most likely to face gubernatorial hesitation. Major cloud providers including AWS, Google, Microsoft, and Oracle have substantial New York data center operations or planned expansions; several have made public statements about their investment commitments to the state. Environmental advocates and union building trades are the moratorium’s most powerful backers; the tech industry is its most organized opponent.

Correction (November 15, 2026): 19 sentences were corrected or removed to match the sources listed on this page.

policy new-york data-centers regulation ai-safety child-safety surveillance
Share

Related Stories

Colorado's AI Law Looms as Washington Wages War on State-Level Regulation

Colorado's comprehensive AI law takes effect June 30, 2026, imposing algorithmic impact assessment requirements on high-risk AI systems — just as the Trump administration intensifies its push to preempt state AI regulation through executive action and litigation. The collision course creates the most complex compliance environment US companies have faced since GDPR.

5 min read

China's Reported $295 Billion AI Buildout Would Shut Out Nvidia and AMD

Beijing is preparing a five-year, 2-trillion-yuan ($295 billion) plan to build a nationwide network of AI data centers. The plan would require at least 80% domestic technology, effectively shutting out Nvidia and AMD, with Huawei standing to gain the most. Including power infrastructure, the total investment could reach at least 5 trillion yuan.

5 min read