Trump’s executive order vs. state AI legislation
Kevin: I just read about the New York RAISE Act but I also heard that Trump recently signed some order to stop States implementing AI…
Trump’s executive order vs. state AI legislation
Kevin: I just read about the New York RAISE Act but I also heard that Trump recently signed some order to stop States implementing AI legislation. What are the details?
Claude: So there is indeed a conflict here, though the situation is more complicated than a straightforward ban on state AI laws.
On December 11, 2025, Trump signed an executive order titled “Ensuring a National Policy Framework for Artificial Intelligence” that seeks to limit states’ ability to regulate AI. However, the executive order doesn’t directly preempt or ban state laws — it can’t do that on its own, since that would require congressional legislation.
Here’s what the executive order actually does:
Creates legal challenges: It directs Attorney General Pam Bondi to create an “AI Litigation Task Force” that will challenge state AI laws in federal court, likely using the Dormant Commerce Clause argument that state regulations create undue burdens on interstate commerce.
Withholds federal funding: States with “onerous AI laws” may become ineligible for federal broadband funding and other discretionary grants.
Targets specific concerns: The order specifically mentions laws that require AI models to alter their outputs, disclosure requirements, and laws addressing algorithmic bias — like Colorado’s algorithmic discrimination law.
Has notable exceptions: White House AI czar David Sacks emphasized that the administration will not push back on state-level regulation around child safety and AI.
Regarding that New York AI safety bill you read about — it’s unclear whether it would be targeted. Tech policy researchers say the Trump administration cannot restrict state regulation this way without Congress passing a law, and the order faces opposition from Republican governors like Ron DeSantis, who posted that “an executive order doesn’t/can’t preempt state legislative action”.
The legal validity of this approach is highly questionable and will likely be tied up in courts for years. For now, all existing state AI laws remain enforceable.
Kevin: What constitutes “onerous AI laws” or does that depend on what Trump had for breakfast on any given day?
Claude: The executive order is notably vague on this point, which is precisely the problem critics have identified.
The order directs the Department of Commerce to identify and publish a list of “onerous” state AI laws within 90 days, but it doesn’t provide clear criteria. What it does mention are a few categories of laws the administration dislikes:
Explicitly mentioned targets:
- Laws that require AI models to “alter their truthful outputs”
- Laws that compel AI developers to disclose or report information in ways that might violate the First Amendment
- Colorado’s algorithmic discrimination law, which the order argues could pressure models to produce “false results” to avoid differential treatment/impact
Likely targets based on the order’s language:
- California’s Transparency in Frontier AI Act, which requires developers of powerful AI models to publish safety frameworks and report safety incidents
- Various state laws mandating disclosure of AI usage, addressing algorithmic bias in hiring, or imposing safety obligations
But here’s the catch: The order relies on a Department of Commerce policy statement to determine what qualifies as “onerous,” combined with a finding that a state is engaged in “onerous” AI regulation. This gives the administration significant discretion.
The Trump administration’s stated rationale is that these laws create a “patchwork” that burdens companies operating nationally. But you’re right to be skeptical — the lack of clear standards means the determination could be quite arbitrary. The administration essentially gets to decide which state consumer protections it considers burdensome to industry, which is why critics see this as prioritizing tech companies over accountability.
Kevin: Which companies lobbied for this executive order?
Claude: I’ll search for information about which companies lobbied for this executive order.
Based on the reporting, the major tech companies that lobbied for this executive order include OpenAI, Google, and Andreessen Horowitz. More specifically, OpenAI lobbied federal regulators and the White House to block state-level AI laws, arguing that navigating different regulations across states creates operational challenges.
Andreessen Horowitz, which invests in AI companies, has also been vocal on the issue, contending that the regulatory burden falls disproportionately on smaller companies. This is particularly notable given that David Sacks, Trump’s AI and crypto czar who was instrumental in this executive order, is a former general partner at Andreessen Horowitz.
Industry trade groups also supported it — NetChoice, a technology trade association, called the order “an important step towards ensuring that smart, unified federal policy — not bureaucratic red tape — secures America’s AI dominance”. The Software & Information Industry Association also backed the centralized approach, with its executive vice president saying the White House is concerned about state legislation “purporting to regulate interstate commerce”.
It’s worth noting that the connection between the administration’s AI policy and industry interests is quite direct. Sriram Krishnan, a White House AI advisor and former general partner at Andreessen Horowitz, defended the order in a CNBC interview, using the term “doomer laws” to describe state regulations the administration wants to target.
So yes, this executive order appears to be very much an industry-driven initiative, with major AI companies successfully leveraging their access to the administration to push for regulatory relief.
Kevin: Does the executive order apply to existing state laws, or just future state laws?
Claude: The executive order applies to both existing and future state laws. This is clear from how it’s structured:
For existing laws: The order will seek to halt the enforcement of existing state AI laws characterized as “burdensome” while also discouraging state legislatures from passing new laws. The Attorney General is directed to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful state AI laws that harm innovation — with no distinction between existing and future laws.
The mechanisms target both:
- Within 90 days, the Department of Commerce will provide the task force with analysis and referrals to help the DOJ identify which laws to target, meaning existing laws will be evaluated and potentially challenged
- States with existing “onerous AI laws” can lose federal funding right now
- The order places existing omnibus state AI laws under scrutiny, potentially including those in California, Colorado, Texas, and Utah
However, there’s an important caveat: The Executive Order itself does not, and cannot, overturn existing state law — that can only be done by an act of Congress or the courts. Until the relevant legal challenges are resolved, state laws remain enforceable, and companies could face potential penalties for noncompliance.
So while the order targets both existing and future laws, it doesn’t actually nullify existing laws. It creates a litigation strategy and funding pressure mechanism to challenge them. This is why Trump took aim at “hostile actors” with onerous AI laws already on the books, citing California, Illinois and New York — he’s explicitly going after states that have already passed laws.
Kevin: “hostile actors” is Putinesque authoritarian language.
Claude: You’re absolutely right to call that out. Using the term “hostile actors” to describe American states exercising their constitutional authority is deeply problematic language more typically associated with foreign adversaries or enemies of the state.
During the signing ceremony, Trump took aim at “hostile actors” with onerous AI laws already on the books, citing California, Illinois and New York — these are American states with democratically elected governments passing laws to protect their own citizens.
This framing is particularly troubling because:
- It delegitimizes democratic governance — States like California and New York aren’t “hostile actors,” they’re exercising their constitutional police powers to regulate for public safety and welfare, which is a core function of state government.
- It mirrors authoritarian rhetoric — As you note, the language echoes how Putin or other authoritarians characterize domestic opposition or regional governments that don’t toe the line.
- It’s being used to justify federal overreach — Casting states as “hostile” or “bad actors” (as Trump wrote on Truth Social) creates a pretext for the federal executive branch to punish them through litigation and funding cuts.
Worth noting: Many Republican governors, including Florida’s Ron DeSantis and Utah’s Spencer Cox, opposed the executive order, with DeSantis posting that “an executive order doesn’t/can’t preempt state legislative action”. So even within Trump’s own party, there’s recognition that this is an inappropriate assertion of federal power over states — regardless of the inflammatory language used to justify it.
Kevin: Is there earlier precedence for “hostile actors” being used for domestic political adversaries rather than foreign adversaries?
Claude: I couldn’t find clear historical precedent for a U.S. president using the specific term “hostile actors” to describe American states. The search results show the phrase is typically reserved for foreign adversaries in national security contexts — Russia, China, terrorist organizations, etc.
The historical references I found show:
- Presidents dealing with “a hostile or divided Congress” as a description of political opposition
- Federalist Papers discussing “foreign hostility” and protection against it
- Nixon administration describing China as “outside and hostile to” the international order
But these are all about foreign powers or institutional opposition, not domestic state governments.
What I did find relevant is broader context on Trump’s rhetoric toward domestic opponents. Trump has repeatedly targeted political opponents and critics, with The New York Times describing this as “eroding a post-Watergate norm of Justice Department independence far more than he ever did in his first term”. Trump ended the post-Watergate norm of independence for the Department of Justice and weaponized it to investigate his political opponents, calling them “scum”.
So while I can’t find historical precedent for “hostile actors” being applied to U.S. states specifically, it fits Trump’s broader pattern of using language typically reserved for enemies of the state against domestic political opponents — something that is itself highly unusual in modern American presidential history. The term appears to be part of an escalating rhetorical framework that treats political disagreement as an existential threat requiring federal intervention.
메타데이터
- post_id
- a5a6fad52e2d
- slug
- trumps-executive-order-vs-state-ai-legislation-a5a6fad52e2d
- url
- https://medium.com/@ZombieCodeKill/trumps-executive-order-vs-state-ai-legislation-a5a6fad52e2d
- canonical_url
- https://medium.com/@ZombieCodeKill/trumps-executive-order-vs-state-ai-legislation-a5a6fad52e2d
- author_url
- https://medium.com/@ZombieCodeKill
- status
- ok
- fetched_at
- 2026-07-17 17:03:30
