The Shot Across the Bow
The Public AI Brief · Issue No. 23
Apologies for the nautical references, I’ve been deep into the Patrick O’Brian series lately. For months, we’ve watched two ships approach each other: states building AI governance frameworks to protect constituents, and a federal government threatening to sweep those efforts aside in the name of innovation. This week, the collision happened. President Trump signed an executive order targeting what the White House calls “cumbersome” state AI regulation, instructing federal agencies to identify which state laws undermine national competitiveness. Legal experts immediately called the action illegal. State leaders and civil rights groups called it dangerous. California called it corruption.
The irony is hard to miss. Just days before Trump’s order, Florida Governor Ron DeSantis proposed an “AI bill of rights” prioritizing citizen protection over Silicon Valley preferences. DeSantis’s Florida House now plans a comprehensive, multi-committee study of AI’s implications, even as the Trump administration moves to block exactly that kind of state-level scrutiny. The Republican coalition on technology policy isn’t just fractured - it’s at war with itself.
Meanwhile, states aren’t backing down. New York Governor Kathy Hochul is playing hardball with legislative leaders over rewriting the RAISE Act. Louisville hired its first Chief AI Officer with a $2 million budget. Alaska is exploring agentic AI for its citizen services portal. Perhaps their CIO, Bill Smith, is just making Alaska the next frontier instead of the last. The executive order may have been signed, but the fight over who governs AI in America is just beginning.
This Week’s Key Developments:
Trump signs executive order targeting state AI regulation; legal experts call it illegal, states call it dangerous
DeSantis breaks with Trump as Florida plans comprehensive AI study while White House moves to block state action
Louisville names first Chief AI Officer with $2M budget as cities continue building capacity despite federal threats
New York plays hardball on AI legislation as Hochul and lawmakers clash over RAISE Act rewrite
Federal
President Trump signed an executive order this week seeking to limit states’ abilities to enact AI-related policy that could be deemed “burdensome.” The order instructs certain federal agencies to identify which state laws undermine federal efforts to help the U.S. lead globally in AI. Experts argue the action is illegal. StateScoop reported state leaders and civil rights groups responding to what they call a “dangerous” order, while California issued a statement declaring Trump’s order “advances corruption, not innovation.”
The order has particular implications for California’s SB 53, which set safety disclosure requirements for companies operating AI models. The law represents exactly the kind of state-level AI oversight the Trump administration now seeks to preempt. The tech industry’s Center for Data Innovation welcomed the order, arguing it “sends the right message on fragmented state laws.” But Federal News Network’s analysis suggests the order “could deepen trust crisis, not solve it,” noting that if the regulatory ecosystem was already messy, a nationwide legal battle over AI and federalism will only intensify the problem.
The White House also instructed agencies to stop using “biased” AI (for my students, that’s a political statement and has no relation to any reality), with the Office of Management and Budget clarifying steps agencies must take to ensure contracted large language models do not produce “woke” outputs. The directive arrives as a bipartisan bill looks to help government hire more AI talent following the exodus of hundreds of thousands of government employees under the administration’s push to shrink the workforce.
Congress established a new House commission to scrutinize AI’s impact on the economy, addressing policy issues like guardrails for AI and its economic, safety and health impacts. Meanwhile, the National Defense Authorization Act includes a directive for DOD to prioritize AI for mental health needs, with the House Armed Services Committee stating “the rate of military suicide is unacceptably high and a new approach is required.”
State
The Trump-DeSantis Split
Florida’s “AI Week” was crashed by the Trump-DeSantis battle over regulation. Just days before Trump’s executive order, DeSantis proposed an “AI bill of rights” marking a sharp departure from the administration’s deregulatory posture. The Florida House now plans a comprehensive, multi-committee study of AI’s implications, examining exactly the kind of state-level policy questions the White House executive order seeks to prevent.
The conflict reveals a deeper fracture in Republican governance philosophy. DeSantis framed his proposal around an “age of darkness and deceit,” arguing citizen protection should take precedence over industry preferences. The Trump administration’s order explicitly prioritizes industry concerns about “50 discordant” state laws hindering competitiveness. One governor sees constituent protection as the primary responsibility; the other sees it as an obstacle to innovation.
New York Plays Hardball
Governor Kathy Hochul is playing a game of chicken with legislative leaders over AI regulations. Hochul wants to rewrite the RAISE Act with language nearly identical to California’s law - a hard no for lawmakers, particularly now that California’s approach faces direct federal challenge. The standoff demonstrates how Trump’s executive order complicates state-level AI policymaking even in blue states that might otherwise align on regulatory approaches.
Hochul isn’t retreating entirely from AI investment. New York committed $40 million to AI training and nuclear energy, pouring resources into clean energy workforce development with aims of expanding the state’s nuclear and AI capabilities. But the governor’s administration is simultaneously running AI-generated advertisements featuring AI-created faces without disclosing the technology to viewers, even as lawmakers push to regulate exactly that kind of AI use in advertising. The disconnect between the state’s regulatory ambitions and its own practices highlights the governance challenges AI creates.
States Building Despite Federal Threats
Route Fifty examined how state Chief AI Officers are defining their roles in real time, noting it remains early days as states stay uncertain about the technology’s future and how to address its impacts. Alaska is considering adding agentic AI modules to the myAlaska app, which residents use for key services, with a recent request for information seeking industry input.
New Hampshire’s Commission on Government Efficiency recommended splitting apart the health department and embracing AI, with the commission’s final report calling for breaking the department into smaller agencies to improve accountability. The recommendation reflects growing recognition that AI adoption requires organizational restructuring, not just technology deployment.
Local
Louisville hired its first Chief AI Officer, with Pamela McKnight coming from the private sector to lead the city’s AI strategy. Officials announced plans to hire a CAIO and build out an AI team earlier this year, powered by a $2 million budget expansion. Louisville’s investment demonstrates cities’ willingness to build governance capacity even as the federal-state regulatory landscape grows more chaotic.
The local government tech sector continues attracting investment. Madison AI raised $3.5 million for its chatbot and AI-backed services to cities, counties and local agencies. The young company’s backers include several government technology veterans along with officials from Nevada, suggesting confidence in municipal AI adoption regardless of federal-state tensions.
Data center tensions persist at the local level. Indiana residents are pushing for a one-year moratorium on large hyperscale data centers in Starke County, with the proposal headed to the County Board of Commissioners for consideration. Google’s plans to reopen Iowa’s nuclear plant to power nearby data centers raise questions about extreme weather threatening reactor safety, particularly after a 2020 storm prematurely shut down the state’s only nuclear facility.
San Jose faces energy concerns as AI’s planned data center boom could nearly triple the city’s energy use, straining California’s grid forecasts and raising fears customers could pay for upgrades if projects never materialize. These local land use and energy decisions remain firmly in municipal hands, one area where Trump’s executive order cannot reach.
Education
Universities Navigate Uncertain Terrain
Texas Christian University committed $10 million to expand AI use through a partnership with Dell, accelerating AI deployment on campus while implementing systems that keep critical data in-house. The private research university’s investment reflects higher education’s push to build AI infrastructure despite regulatory uncertainty.
Two years into the NAIRR pilot, shared infrastructure is boosting AI innovation. The National Artificial Intelligence Research Resource pilot connects researchers, educators and industry partners, providing shared computing power, AI tools and educational support for pushing boundaries with the technology.
Federal Scrutiny of EdTech
Federal officials are floating tying kids’ screen time to school subsidies, with NTIA Administrator Arielle Roth stating the agency will study whether schools are too reliant on educational technology and if spending has resulted in bad outcomes for students. The proposal represents federal willingness to intervene in educational technology adoption even as the administration moves to block state AI regulation in other contexts.
Indiana lawmakers are taking up a bill expanding cellphone bans to the entire school day, building on existing state law restricting student use during instructional time. The “away for the day” approach shows states addressing technology’s educational impacts through straightforward restrictions rather than complex AI governance frameworks.
Key Insights for Practitioners
Legal challenges will define AI governance for years, not months: Trump’s executive order won’t settle the federal-state question - it opens a legal battle that will work through courts while agencies continue deploying AI systems. State and local governments should build governance capacity assuming regulatory uncertainty persists.
Action: Establish AI governance frameworks that can adapt to changing federal-state dynamics. Focus on internal policies, procurement standards, and workforce development that provide value regardless of how preemption battles resolve.
Republican fracture creates state-level opportunity: DeSantis’s break with Trump on AI regulation reveals Republican governors aren’t uniformly aligned with federal deregulation. States led by governors prioritizing constituent protection over industry preferences may find unexpected allies across party lines.
Action: Monitor which Republican governors follow DeSantis’s model versus Trump’s approach. Build coalitions based on shared governance priorities rather than partisan alignment.
Local governments hold leverage federal orders can’t touch: Cities and counties control land use permitting, energy infrastructure decisions, and procurement choices. Trump’s executive order targets state AI laws, not municipal authority over data centers, energy use, or vendor selection.
Action: Assert local control over AI infrastructure questions. Use land use authority, energy planning, and procurement standards to shape AI deployment in your community regardless of federal-state battles.
What I’m watching: Which states sue to challenge the executive order, and whether Republican attorneys general join the legal fight. The Trump-DeSantis split suggests at least some Republican states might defend their AI regulatory authority rather than defer to federal preemption. If red states and blue states both challenge the order, the legal landscape gets significantly more complicated for the White House.
The federal government just fired a shot across the bow. How are states responding in your jurisdiction? Are local leaders building AI capacity or waiting for federal-state clarity? Share your observations in the comments.