8 AI policy moves you should know about — week of July 14, 2026
New York became the first state to impose a statewide data center moratorium; Prince George's County set the longest local moratorium in the country; and Tennessee emerged as the new epicenter of local data center resistance.
AI Laws USA tracks every U.S. federal, state, county, and city AI law on a single live map at ailawsusa.com. The week of July 14, 2026 saw the most significant escalation in data center regulation since local moratoriums began proliferating in 2025: for the first time, a governor stepped in. Governor Hochul's executive order pausing large data center permitting in New York was the biggest individual policy move of the year so far. Meanwhile, Prince George's County, Maryland set a two-year moratorium — the longest county-level pause in the country — and Tennessee became the nation's most active local battleground. Here are the eight moves that mattered most.
-
MORATORIUM
1. New York became the first state in America to impose a statewide data center moratorium
New York (state — Governor Hochul Executive Order) · Signed July 14, 2026 — effective immediately; moratorium lasts approximately one yearGovernor Kathy Hochul signed an Executive Order on July 14, 2026 pausing state permitting for new data centers planning to draw more than 50 megawatts of electricity — the first statewide data center moratorium issued by any U.S. governor. The order directs the Public Service Commission to conduct a generic environmental and energy review of data center impacts on the state grid, and directs Empire State Development to create a 'Community Interest Framework' within 60 days so local governments have tools to negotiate with data center developers. The EO is separate from the pending legislative bill S10642 (which Hochul has not yet signed or vetoed); it accomplishes a moratorium through the Governor's authority over state permitting systems. New York — home to some of the most active AI infrastructure corridors in the Northeast — is now the first state to say: pause.
-
MORATORIUM
2. Prince George's County, Maryland imposed the longest county-level data center moratorium in the country — two years
Prince George's County, MD (county — County Council) · Adopted July 7–8, 2026 — in effect through approximately July 2028The Prince George's County Council voted on July 7–8, 2026 to impose a two-year moratorium on all new hyperscale data center development in the county — the longest county-level data center moratorium imposed anywhere in the United States. The action was catalyzed by the proposed conversion of the former Landover Mall site into a large data center complex, which drew intense community opposition. The moratorium allows the Council to draft a comprehensive zoning framework setting standards for data center siting and construction before any new permits are issued. The moratorium may be lifted early if a permanent zoning ordinance is adopted. One Council member voted no; three abstained. PGC is one of the most active data center markets in the Mid-Atlantic.
-
MORATORIUM
3. Tennessee became the nation's most active local battleground for data center restrictions
Hamblen County, TN (county); Shelbyville, TN (city) · Hamblen County: enacted July 6, 2026; Shelbyville: enacted July 9, 2026Two more Tennessee communities enacted data center moratoriums this week, bringing Tennessee's local restriction wave to one of the densest concentrations in any single state. Hamblen County commissioners voted July 6 to enact a one-year AI Datacenter Moratorium — the county's official government website uses that name, making it one of the few local jurisdictions to explicitly invoke 'AI' rather than 'data centers' in the title. The City of Shelbyville added a unique twist: its moratorium covers both data centers and cryptocurrency mining plants, reflecting community concerns about electricity demand from any large-scale computing facility. Both join Coffee County, Cedar Hill, McMinnville, and Somerville — all of which acted in June and July 2026.
-
PROPOSED
4. Walker County, Georgia passed a 30-day bridge moratorium — and Hamilton County, Tennessee has a vote scheduled for July 15
Walker County, GA (county); Hamilton County, TN (county — vote July 15) · Walker County: enacted July 10, 2026; Hamilton County vote: scheduled July 15, 2026Walker County, Georgia — which borders Hamilton County, Tennessee — passed a 30-day emergency data center moratorium on July 10, 2026, providing time to draft permanent zoning language. The very short duration is characteristic of 'bridge' moratoriums: the county pauses new applications while commissioners draft a permanent ordinance. Meanwhile, Hamilton County, Tennessee (home to Chattanooga) is scheduled to vote on its own data center moratorium on July 15, 2026. If Hamilton County acts, it would join Walker County just across the state line — creating a rare cross-state pair of simultaneous restrictions in the same regional market.
-
PROPOSED
5. Hawaii's AI disclosure and chatbot safety law is expected to become law by July 15 — with or without the Governor's signature
Hawaii (state — SB 3001) · Enrolled; Governor Green must sign, veto, or allow to lapse by July 15, 2026 — signature not yet confirmedHawaii SB 3001 (CD1) — the AI Disclosure and Safety Act — is enrolled and awaiting Governor Josh Green's decision, which must come by July 15, 2026. Governor Green's intent-to-veto list published June 26–27 does not include SB 3001, making it likely the bill either receives a signature or becomes law via inaction. The bill requires AI companion chatbots to disclose their AI nature, bans companion chatbots from charging fees or collecting payment information from minors, and directs the state to publish a public register of high-risk AI systems. If enacted, Hawaii would join Oregon and Washington in the growing cluster of Pacific states with specific AI-companion protections.
-
PROPOSED
6. The FTC proposed that some state AI laws may be preempted by federal consumer protection law
United States (federal — FTC Policy Statement) · Published in Federal Register July 7, 2026 — public comments due July 31, 2026The Federal Trade Commission published a proposed policy statement in the Federal Register on July 7, 2026 warning that AI companies steering outputs toward undisclosed ideological objectives may violate Section 5 of the FTC Act. Issued pursuant to a Trump Executive Order (EO 14365), the statement makes a potentially far-reaching argument: state laws that require AI systems to alter their outputs — including some state AI bias and accuracy mandates — may themselves conflict with federal consumer protection law. The preemption argument is aimed at state AI regulations, including provisions in Colorado's SB 26-189. The statement is proposed, not final, and has no immediate enforcement effect. But it signals the direction of a deregulatory federal posture toward state AI oversight. Comments are due July 31, 2026.
-
NEW LAW
7. California's healthcare AI misrepresentation ban went into effect January 1 — and is newly indexed this week
California (state — AB 489) · Signed October 11, 2025 — in effect since January 1, 2026California AB 489, sponsored by the California Medical Association and signed by Governor Newsom on October 11, 2025, prohibits AI systems from displaying post-nominal letters (M.D., R.N., Ph.D., etc.), icons, or phrases that imply a user is interacting with a licensed healthcare professional unless actual licensed oversight exists. It also bars marketing language — 'doctor-level,' 'clinician-guided,' 'expert-backed' — without genuine licensed support. The law expands the enforcement authority of California's professional licensing boards (Medical Board, Board of Registered Nursing, Psychology Board), with each misleading representation treated as a separate offense. The law was in effect before this sweep but is newly indexed in the AI Laws USA dataset this week. It is a companion to AB 3030 (healthcare AI notification in clinical-note generation) and SB 1120 (the Physicians Make Decisions Act), collectively giving California the most detailed healthcare AI disclosure regime of any state.
-
NEW LAW
8. Oregon's AI companion chatbot law — the first with a direct private right of action — is newly indexed and takes effect January 1, 2027
Oregon (state — SB 1546) · Signed March 31, 2026 — effective January 1, 2027Oregon SB 1546, the AI Companion Chatbot Safety Act (signed March 31, 2026, effective January 1, 2027), is the first U.S. chatbot law to give any harmed person a direct right to sue: $1,000 in statutory damages per violation, greater actual damages, injunctive relief, and attorney fees. The law requires AI companions — systems designed to simulate a sustained human-like relationship while retaining context across sessions — to disclose their AI nature at the start of every interaction, repeat that disclosure every three hours for adults and every hour for minors, implement crisis protocols for suicidal ideation, and ban manipulative engagement mechanics (variable-ratio reinforcement, unpredictable rewards) for minor users. Oregon's private right of action makes it meaningfully stronger than Washington HB 2225, which enforces only through the Attorney General. The law takes effect in fewer than 200 days.
The dominant theme this week is geography: data center resistance is no longer a coastal phenomenon. Tennessee and Georgia communities are acting faster than any other region in the country, and a New York governor's pen just raised the stakes to the statewide level for the first time anywhere in the U.S. Simultaneously, the FTC's preemption signal introduced a new tension — the question of whether federal consumer protection law will be used to limit, rather than expand, state AI oversight. Both threads will move fast. Track all of them, plus every law already in effect for any U.S. address, at ailawsusa.com. New posts every Monday.