The AI Policy Podcast

Episode: Trump Signs EO Targeting State AI Laws While Meta Showcases Risks of Weak Tech Regulation
Date: December 18, 2025
Host: Gregory C. Allen (Greg), Senior Adviser, Wadhwani AI Centers at CSIS
Co-Host: Sadie McCullough

Episode Overview

This episode unpacks President Trump’s recent executive order (EO) aimed at curtailing state-level AI regulation, seeking to establish a federal AI regulatory standard, and examines bombshell reporting from Reuters about Meta’s (Facebook’s parent company) alleged facilitation of scams under light regulation. The discussion covers details of the EO, the reactions from supporters and critics, potential legal battles, and draws lessons from Meta’s actions on the necessity and challenges of tech regulation.

Key Discussion Points and Insights

1. President Trump’s Executive Order on State AI Laws

Background and Purpose ([00:10]–[01:52])

• Trump signed an EO on December 11 intended to preempt state AI regulatory laws in favor of federal standards, citing interstate commerce as justification.
• The final text builds on previous drafts, with mechanisms to both fight state regulation and introduce federal oversight.

Main Mechanisms of the EO ([01:52]–[06:54])

• Department of Commerce tasked with reviewing and challenging existing state laws deemed “overly burdensome.”
• AI Litigation Task Force: Department of Justice team to file lawsuits against such state laws.
• Federal Funding Leverage: Potential restrictions on broadband, equity, and research funds to states, mirroring strategies used in past federalizing efforts.
  • Greg notes this approach:

    “It's a way that the federal government has in the past tried to federalize certain types of legal regimes... These ones, they're saying, are sort of tied to AI, which... is meant to make it more likely to survive court review.” [04:18]

• FTC Preemption: Clarifying circumstances under which state laws requiring output alterations from AI models are preempted by the FTC Act—they target “ideologically motivated outputs,” with specific reference to events like the Gemini AI launch debacle ([04:57]).
• Federal Standards Introduction:
  • FCC tasked to consider national reporting/disclosure standards for AI within 90 days.
  • The White House’s AI czar (David Sachs) and OSTP director (Michael Kratzios) directed to propose uniform federal AI policy recommendations to Congress.

Key Concessions and Changes ([06:54]–[08:38])

• Purpose section additions: federal framework must address child protection, censorship, copyright, and community safety.
• Carveout: The EO explicitly will not seek to preempt state AI laws related to child safety.

Support for the EO ([08:38]–[11:26])

• Key supporters: David Sachs (AI czar), Sriram Krishnan, Senator Ted Cruz.
  • Sachs (on X):

    "This is not AI amnesty or AI moratorium. It is an attempt to settle a question of jurisdiction." [09:13]

  • Ted Cruz (on X):

    "President Trump took an important step today to promote American leadership in AI. We don't want China's values... We want American values of free speech, individual liberty and respect for the individual." [09:32]

• Dean Ball (former White House OSTP, now at a think tank):

  "On its own, the EO cannot create a nationwide preemptive framework for that. We will need a new law passed by Congress." [10:43]

• Strategy theme: EO sets up legal and administrative challenges but still requires Congressional action for lasting federal standards.

Bipartisan Opposition and Key Arguments ([11:26]–[14:46])

• Opposition comes from both parties:
  • Republican: Gov. Ron DeSantis (Florida) objects on state rights and process:

    “An executive order... can't preempt state legislative action. Congress could theoretically... but hasn't proposed any coherent regulatory scheme.”

    • “This would be an AI amnesty.” [12:27]
  • Democratic: Gov. Gavin Newsom (California) opposes EO, recently passed a state AI safety law (SB53).
  • Representative Doug Fiefia (Utah, Republican):

    “Trump’s new order is an overreaching act that fundamentally disregards the 10th Amendment...” [13:28]

• General sentiment: Most critics favor a national framework developed through Congress, not via executive order.

Expected Legal Battles ([14:46]–[17:50])

• California’s SB53 and similar state laws likely to be targeted; state leaders (e.g., Sen. Scott Wiener) prepare for court fights.
  • Scott Wiener:

    “If the Trump administration tries to enforce this ridiculous order, we will see them in court.” [15:24]

• Even sympathetic commentators like Dean Ball estimate low (30–35%) likelihood the EO survives legal challenges.

White House's Outreach to Republican Skeptics ([17:50]–[20:12])

• Direct engagement with senators like Marsha Blackburn to address concerns that the order threatens states’ rights.
• White House argues they’re leaving leeway on certain harms (esp. child safety) and focusing on pre-market requirements rather than bans on harm regulation.

Anticipated Effects and Regulatory Uncertainty ([20:12]–[21:54])

• Court battles are certain, and meanwhile, blue states will likely “ignore” the order, deepening regulatory uncertainty.
  • Shaquille Hashim (Transformer editor):

    “It’s legally messy, possibly unconstitutional, and will probably just be ignored by blue states... That will have the effect of creating less regulatory certainty rather than more.” [20:34]

• Bottom line:

  “Executive action alone is not enough... All sides agree, but little progress is being made in Congress.” [10:51]

2. Meta and the Dangers of Under-Regulation

Reuters’ Bombshell Reporting on Meta ([22:39]–[38:39])

• Reuters accessed internal Meta documents revealing:
  • Meta projected ~$16bn (10% of revenues) from advertising scams and banned goods in 2024.
  • Meta’s own staff estimate: 1/3rd of all successful scams in the US involved Meta platforms; UK regulator found 54% of payment scams linked to Meta.
  • Meta actively restricted antifraud teams from taking measures that would cost more than 0.15% of total revenue.
  • User reports of fraudulent/scam ads are ignored or incorrectly rejected 96% of the time.
  • Meta charges higher rates (“scam tax”) for ads suspected to be fraudulent.
  • Repeat offenders: Small advertisers must be flagged 8 times; “high value” advertisers can get 500+ strikes before being banned.

  • “15 billion scam ads presented to users daily, plus 22 billion organic scam attempts.” [38:49]

• Senators' response:
  • Senators Blumenthal and Hawley’s open letter ([28:09]):

    “By its own assessment, 10% of Meta’s annual revenue ... was generated from scam advertisements, a windfall of $16bn in one year alone. ... Meta was responsible for more than $50bn in consumer loss last year.” “Meta had specific revenue guardrails that blocked its anti-fraud team from taking action that would cost Meta more than ... 0.15% of total revenue.” “Meta reportedly charges higher rates for ads that it suspects might be fraudulent, in effect imposing a scam tax...” [29:42] “These scams have been allowed to take over Facebook and Instagram as Meta has drastically cut its safety staff, including for FTC-mandated reviews, even as it dumps unimaginable sums into its generative AI projects ...” [31:06] “Meta will not act on this rampant fraud and criminality unless the FTC and SEC put an end to it.”

International Aspect and Further Revelations ([40:44]–[45:04])

• China is the origin of roughly 1/4th of scam ads worldwide; Meta benefits from these despite China banning access for its own citizens.
  • Meta tried (successfully) to reduce scam ads, but Mark Zuckerberg shut down those efforts when revenues fell.

  • “After Zuckerberg’s input, the documents show Meta disbanded its China-focused anti-scam team. ... By mid-2025, banned ads climbed up to about 16% of Meta’s China revenue.” [44:00]

  • Greg: “They know the harm is being done. They know how to stop it. ... But when it worked, they didn’t like the results and so they shut down the operation to stop it.” [44:43]

Lessons for AI and Tech Regulation ([45:21]–[55:01])

• Regulatory fines are simply a cost of doing business; Meta internally acknowledged that profits from scams exceed any anticipated fines.

  • “[Meta’s] revenue from this pattern of behavior exceeds the expected fines. So the fines are just the cost of doing business.” [47:00]

• Weak regulation, even if the actions are illegal, fails to deter harmful conduct.
• The need for transparency:
  • “If the government does not have a mechanism for visibility into these platforms, the window of opportunity for detecting this and shutting it down is going to be pretty tough.” [52:42]
• Drawing parallels to social media’s regulatory failures as a warning for AI:
  • Demis Hassabis (Google DeepMind):

    “We should learn the lessons from social media where this attitude of maybe move fast and break things went ahead of the understanding of what consequent ... effects were going to be.” [53:23]

  • Senator Blumenthal:

    "We had the same choice when we face social media, we failed to seize that moment. ... Now we have the obligation to do it on AI before the threats and the risks become real." [53:47]

• Greg’s take:
  • Existing laws aren’t enough; technological change enables both sophisticated harm and greater possibility for intervention, but platforms won’t act without tougher regulations or transparency.
  • “Don’t take Meta’s public statements at their word ... there’s gotta be some degree of transparency that allows for a ‘trust but verify’ approach when the stakes are this high.” [54:57]

Notable Quotes & Memorable Moments

• Greg on the EO’s legal risk:

  “I think the administration has a 30 to 35% chance of this working legally, which I think is so interesting.” [16:32]

• Greg on court strategy:

  “The Trump administration can get some of what they want just by raising the specter of court... Because that can, you know, be in the background of state legislatures’ minds as they’re debating, well, is this really something that we want to go to the mattresses over?” [16:50]

• Greg on Meta’s internal risk math:

  “They know they’re going to lose in court... but we don’t care because our revenue from this pattern of behavior exceeds the expected fines.” [47:00]

• Greg on transparency and tech regulation:

  “While we want to assume the best of intent... there’s gotta be some degree of transparency that allows for a trust but verify approach when the stakes are this high.” [54:57]

Key Timestamps

|Timestamp|Segment / Topic| |---|---| |00:10 – 06:54| Executive Order: Overview, mechanisms, changes from draft| |08:38 – 11:26| Supporters and their arguments (Sachs, Cruz, Ball)| |11:26 – 14:46| Opposition: Bipartisan concerns, states’ rights| |14:46 – 17:50| Fate of state bills, expected legal challenges| |20:23 – 21:54| Impact: Regulatory uncertainty| |22:39 – 34:01| Meta: Introduction, highlights from Reuters, Blumenthal & Hawley letter| |38:49 – 45:04| Meta’s response, China connection, further explosive revelations| |45:21 – 55:01| Lessons for AI policy, deterrence failure, trust but verify| |53:23 – 54:57| Lessons drawn for AI regulatory frameworks

Final Thoughts and Takeaways

• Courts will decide the fate of Trump’s preemption strategy, but the real path to consistent regulation lies through Congress.
• Meta’s case starkly demonstrates that voluntary standards and low-penalty regimes will not prevent or stop massive, even knowingly facilitated, harms.
• For AI and tech regulation, the lesson is clear: transparency, enforceable standards, and adequate deterrence are vital—otherwise, economic and social harms will persist at scale.
• Concluding note:

  “If there’s one thing Meta’s internal documents say incredibly persuasively, it’s don’t take Meta’s public statements at their word...” [54:57]

Recommended reading: Reuters’ entire investigative series on Meta for a deeper, first-hand view of the evidence and implications discussed in this episode.