The Information Machine

US AI Regulation: Federal Retreat vs. State Intervention

closed · v32 · 2026-07-08 · 433 items · history

What's new in v32

No new themes this pass. The new items are additional sourcing for already-established stories — further reporting on sustained US pressure on Meta to join the review system [15][16][17], confirmation that Anthropic's models received clearance after the June 30 export control lift [10], and additional coverage of GPT-5.6 Sol's restricted launch [22]. These items are incorporated into relevant perspectives and timeline entries but introduce no new voices, events, or substantive positions. The Anthropic export control search is retired as that story has resolved and the background is sufficient.

What

The Trump administration is running AI oversight through executive action rather than legislation — a June 3 executive order established a nominally voluntary 30-day pre-release review with NSA oversight [1][3], while in practice the administration has imposed and then lifted export controls on Anthropic [7][9], secured a government-gated launch of OpenAI's GPT-5.6 Sol [20][21], and continued pressing Meta — the only major US AI lab still outside the review system — to join [15][16][17]. Outgoing White House AI adviser Sriram Krishnan confirmed on July 4 that this non-regulatory posture is permanent [6]. State-level enforcement advances independently, with Illinois enacting mandatory auditing [31] and a 42-state AG coalition subpoenaing OpenAI [34].

Why it matters

The administration is exercising de facto control over which frontier AI models reach which users through tools with no clear statutory basis, and a named insider has confirmed no formal regulatory structure will replace them. Meta's continued refusal is the clearest test of whether executive pressure alone can make a voluntary framework universal.

Open questions

  • Will Meta eventually comply with administration review pressure [15][16][17], and does the administration have any enforcement mechanism beyond public statements if it refuses outright?

  • What prompted the White House to lift Anthropic's export control on June 30 [9][10] — did Anthropic make concessions, or was the ban temporary leverage that served its purpose?

  • Krishnan warned the US lacks a leading American open-weight model while China has competitive ones [6] — does the administration's access-gating approach make that gap worse, and will it take separate action to address it?

  • Will the 42-state AG coalition's subpoena [34] and Florida's personal suit against Altman [35] converge on coordinated settlement terms or remain separate enforcement tracks?

Narrative

The Trump administration has built AI governance through executive action. Its June 3, 2026 executive order established a voluntary 30-day pre-release review for frontier AI models with NSA oversight and classified compliance thresholds, explicitly disclaiming binding requirements [1][2][3]. That narrow form reflects direct lobbying by Elon Musk, Mark Zuckerberg, and White House AI czar David Sacks, who opposed binding requirements and shortened the review window from 90 days to 30 [4][5]. Outgoing White House AI adviser Sriram Krishnan confirmed in a July 4 FT interview that this posture is settled policy: Trump will never support a broad US AI regulatory or licensing body, and companies, chipmakers, security firms, and intelligence agencies are expected to manage AI risk themselves [6].

In practice, the administration has gone further than the stated voluntary terms. It forced Anthropic to disable Fable-class models for all foreign nationals on June 14 [7][8], then reversed that export control on June 30 [9][10]. It moved the DOJ to protect xAI's Mississippi data center as national security infrastructure [11] and pressed Meta — the only major US AI lab still outside the review system — to join; the White House publicly expressed confidence an agreement was imminent [12], but as of early July Meta has not complied [13][14][15][16][17]. The clearest extension of executive control came with GPT-5.6 Sol: at the administration's request, OpenAI released the model as a government-gated controlled preview with customer-by-customer approval, citing the model's autonomous offensive cyber capabilities [18][19][20][21][22]. Critics describe this as government-imposed 'AI feudalism' [23] and warn that access restrictions may push developers toward Chinese and open-weight alternatives [24][25].

The argument over oversight has split between domestic and international frames. Anthropic CEO Dario Amodei called for mandatory pre-release testing and government-backed independent auditing [26][27]; White House AI czar David Sacks accused Anthropic of 'a sophisticated regulatory capture strategy based on fear-mongering' [28] and VP Vance echoed that publicly [29]. Peter Thiel extended the anti-regulation argument internationally: he accused Pope Leo XIV of unintentionally acting as a 'Chinese communist agent' by advocating international AI oversight, arguing that global AI rules would bind the United States faster than China, making restraint equivalent to strategic surrender [30]. Krishnan raised a related but distinct concern — that China has competitive open-weight models while the US lacks a leading American open-weight alternative — and warned that Americans need a visible share of AI's economic gains or public resentment will keep growing [6].

State governments are advancing binding AI regulation without federal coordination. Illinois enacted SB 315 requiring annual third-party audits, 72-hour incident reporting, and whistleblower protections [31]. California is advancing SB 947 and SB 951 requiring human oversight of automated employer discipline and advance notice of AI-driven layoffs [32][33]. New York AG Letitia James, on behalf of a 42-state coalition, subpoenaed OpenAI over consumer practices, data handling, treatment of minors, and chatbot sycophancy [34]; Florida expanded its lawsuit to 83 pages naming CEO Sam Altman personally [35]. Trump's preemption executive order challenged state AI laws on constitutional grounds [36], but multiple law firms found its authority limited absent congressional action [37].

Timeline

  • 2026-05: Trump signs preemption EO challenging state AI laws; multiple law firms find constitutional authority limited absent congressional action [36][37]
  • 2026-05: California Senate advances SB 947 requiring human oversight of automated employer discipline and SB 951 requiring 90-day advance notice of AI-driven layoffs [33][70][32][71]
  • 2026-05-21: Trump cancels AI security EO after Sacks, Musk, and Zuckerberg lobby against binding pre-deployment review requirements [4][5][81]
  • 2026-05-28: Illinois Governor Pritzker signs SB 315 requiring mandatory third-party auditing, 72-hour critical incident reporting, and whistleblower protections [72][73][31]
  • 2026-06-01: Florida becomes first state to sue OpenAI and Sam Altman over ChatGPT's alleged role in violent crimes; AG Uthmeier opens a separate criminal investigation [74][75][76]
  • 2026-06-03: Trump signs revised AI executive order establishing a voluntary 30-day pre-release review for frontier models with NSA oversight and classified compliance thresholds [1][2][38][3]
  • 2026-06-04: Build American AI — backed by OpenAI, a16z, and Palantir — acknowledged running false-flag social media accounts; WIRED reports the PAC paid TikTok influencers to fear-monger about Chinese AI [52][53][82][83]
  • 2026-06-10: Anthropic CEO Dario Amodei publishes blog calling for mandatory pre-release testing and government-backed independent auditing [26][27]
  • 2026-06-12: New York AG Letitia James serves OpenAI with a subpoena on behalf of a 42-state coalition targeting advertising practices, data handling, treatment of minors, and chatbot sycophancy [34]
  • 2026-06-14: David Sacks publicly accuses Anthropic of 'regulatory capture strategy based on fear-mongering'; US government forces Anthropic to disable Fable-class AI models for all foreign nationals [28][7][8]
  • 2026-06-15: Florida expands its lawsuit against OpenAI to 83 pages naming CEO Sam Altman personally, separate from the ongoing criminal probe [35][77]
  • 2026-06-18: DOJ moves to protect xAI's Mississippi AI data center, treating high-speed AI compute as national security infrastructure [11]
  • 2026-06-25: US government asks OpenAI to release GPT-5.6 as a controlled preview with customer-by-customer government approval, citing cybersecurity concerns; White House says Meta agreement is imminent [18][12]
  • 2026-06-27: GPT-5.6 Sol launches as a government-gated controlled preview; OpenAI publishes a system card under its Deployment Safety Hub [40][55][56][84][20][21][22]
  • 2026-06-30: White House lifts export control on Anthropic, reversing the June 14 ban blocking Fable-class models for foreign nationals; no public explanation given [9][10]
  • 2026-07-03: Peter Thiel accuses Pope Leo XIV of unintentionally acting as a 'Chinese communist agent' by advocating international AI regulation, arguing global rules bind the US faster than China [30]
  • 2026-07-04: Outgoing White House AI adviser Sriram Krishnan confirms Trump will never support a broad US AI regulatory body; warns the US lacks a leading American open-weight model while China has competitive ones [6]

Perspectives

Trump administration (federal)

Signed a voluntary AI EO disclaiming binding requirements; simultaneously imposed and lifted export controls on Anthropic, secured a government-gated Sol launch, and pressed Meta to join the review system; outgoing adviser Krishnan confirmed no formal regulatory body will ever be established [6].

Evolution: Krishnan's July 4 statement makes the anti-regulatory posture explicit: the coercive actions taken outside the EO's stated terms are the policy, not anomalies.

[36][4][1][2][38][3][11][7][39][18][12][40][9][14][6]

Tech accelerationist faction (Sacks, Musk, Zuckerberg, Vance)

Opposes domestic AI oversight mechanisms; Sacks accuses Anthropic of regulatory capture through fear-mongering; VP Vance echoed that framing publicly.

Evolution: Consistent in opposing binding domestic requirements; their influence is visible in the June 3 EO's narrow voluntary terms, though the administration's subsequent coercive actions sit in tension with the stated accelerationist position.

[4][5][41][28][29][42]

Peter Thiel

Argues that any international AI regulation hands a strategic advantage to China, since rules constrain the US faster than they constrain China, making restraint equivalent to surrender [30].

Evolution: Extends the domestic deregulatory argument into an explicitly geopolitical frame opposing international governance.

[30]

Anthropic / Dario Amodei

Calls for mandatory pre-release testing and government-backed independent auditing; disputed Sacks's 'fear-mongering' characterization while claiming alignment with Trump administration priorities.

Evolution: Export controls on Fable-class models were imposed June 14 and lifted June 30; models have since received clearance [10], but whether Anthropic made concessions or the ban was temporary leverage remains unknown.

[27][26][43][44][45][7][46][39][8][9][10]

OpenAI

Publicly advocates for federal governance and shared evaluation standards; complied with the government's request to launch GPT-5.6 Sol as a controlled, government-gated preview [20][22].

Evolution: The Sol launch represents compliance with executive access control beyond pre-release review; commentary argues this aligns government revenue interests with OpenAI's commercial success, complicating the case for independent oversight [47].

[48][49][50][51][52][53][54][34][18][40][55][56][20][47][22]

Meta

Refusing to join the voluntary pre-release review system despite sustained administration pressure; confirmed holding out as of early July [15][16][17].

Evolution: The White House expressed public confidence an agreement was imminent, but multiple reports through July confirm Meta has not complied.

[57][58][59][60][61][62][63][64][12][65][66][67][13][14][68][69][15][16][17]

State legislators (Illinois + California)

Advancing binding AI safety and employment law independent of federal coordination; Illinois SB 315 enacted; California SB 947 and SB 951 advancing.

Evolution: Consistent; operating without federal coordination and not slowed by the Trump preemption EO.

[33][70][32][71][72][73][31]

State AGs (42-state coalition + Florida litigation)

Pursuing AI accountability through civil and criminal enforcement; the 42-state coalition subpoenaed OpenAI over consumer practices; Florida expanded its suit to 83 pages naming Altman personally.

Evolution: Expanded from Florida's single-state lawsuit to a coordinated 42-state investigation without federal coordination.

[74][75][76][34][35][77][78][79][80]

Tensions

  • Mandatory oversight (Amodei) vs. accelerationist and geopolitical opposition (Sacks, Vance, Thiel): Amodei calls for mandatory pre-release testing and independent auditing [26]; Sacks and Vance argue safety advocates manufacture fear for commercial reasons [28][29]; Thiel contends that global AI regulation binds the US faster than China [30] — yet the administration simultaneously imposed and lifted export controls on Anthropic within two weeks [7][9]. [26][28][29][30][7][9]
  • Federal preemption preference vs. state enforcement reality: Trump's preemption EO lacks constitutional authority absent congressional action [36][37]; binding Illinois law [31] and a 42-state coalition subpoena [34] advance without federal coordination. [36][37][31][34]
  • National security access control vs. risk of accelerating competitor adoption: the government gated GPT-5.6 Sol behind customer-by-customer approval citing cybersecurity [18][20][22], but critics argue this pushes developers toward Chinese and open-weight alternatives [24][25]; Krishnan separately warned the US already lacks a leading American open-weight model while China has competitive ones [6]. [18][20][22][24][25][6]
  • 'Voluntary' review framework vs. coercive application: the administration labeled the review system voluntary [1][3] but imposed and reversed export controls on Anthropic [7][9], pressed Meta to join [12][15][16], and secured a government-gated Sol launch [40][55] — none of which is contemplated in the stated voluntary framework. [1][3][7][9][12][15][16][40][55]
  • OpenAI's governance advocacy vs. its PAC's political operations and legal exposure: OpenAI calls for democratic governance and shared evaluation standards [49] while Build American AI admitted running sockpuppet accounts [54] and the 42-state subpoena targets the consumer practices OpenAI presents as responsible [34]. [49][52][54][34]
  • Government as access gatekeeper vs. government as financial stakeholder: routing GPT-5.6 Sol's commercial access through government approval turns the state financially long on OpenAI's success [47][20], creating a structural conflict of interest that undermines independent oversight. [40][20][47]

Status: active and growing

Sources

  1. [1] Trump Signs Executive Order For AI Testing Prior To Frontier Model Releases — Zvi's AI Roundups (2026-06-03)
  2. [2] President Trump signs executive order to review advanced AI models. — Rohan Paul Twitter (2026-06-03)
  3. [3] Executive Order Creates Voluntary Regulatory Regime of Frontier AI ... — reactive:us-ai-policy-regulation
  4. [4] Trump abruptly cancels EO signing event after top AI firm CEOs declined to go — Ars Technica AI (2026-05-22)
  5. [5] AI #170: Lack of Executive Order — Zvi's AI Roundups (2026-05-28)
  6. [6] FT: Trump will never support a US AI regulator, says outgoing adviser Sriram Krishnan. — Rohan Paul Twitter (2026-07-04)
  7. [7] The US government just forced Anthropic to disable its most advanced AI models for all foreign nationals — and the compa... — reactive:us-ai-policy-regulation (2026-06-14)
  8. [8] David Sacks on the US government's shutdown of Anthropic's Fable ... — reactive:us-ai-policy-regulation
  9. [9] White House lifts export control on Anthropic that froze its most ... — reactive:us-ai-policy-regulation
  10. [10] From Anthropic's new AI models finally getting clearance from US ... — reactive:us-ai-policy-regulation
  11. [11] Large AI campuses are indeed strategic infrastructure of a nation. — Rohan Paul Twitter (2026-06-18)
  12. [12] 'We hope to sign the agreement soon': White House calls on Meta to submit AI models for review, citing abilities and vulnerabilities evaluation | TechRadar — reactive:us-ai-policy-regulation
  13. [13] The US is pressing Meta to let it review its AI, and Meta is the last holdout — reactive:us-ai-policy-regulation
  14. [14] U.S. Presses Meta to Agree to A.I. Reviews - The New York Times — reactive:us-ai-policy-regulation
  15. [15] US presses Meta to agree to AI reviews as security ... — reactive:us-ai-policy-regulation
  16. [16] US presses Meta to agree to AI reviews as security ... — reactive:us-ai-policy-regulation
  17. [17] US presses Meta to agree to AI reviews as security concerns rise — reactive:us-ai-policy-regulation
  18. [18] The Information: The US government is asking OpenAI to slow GPT-5.6 into a controlled preview instead of releasing it br… — Rohan Paul Twitter (2026-06-25)
  19. [19] OpenAI Launches Limited Preview of GPT-5.6 at US Government's ... — reactive:gpt-56-launch-government-access
  20. [20] OpenAI Previews GPT-5.6 Sol With Restricted Access and Stronger ... — reactive:us-ai-policy-regulation
  21. [21] OpenAI Limits Access to New Models, Citing Government Security ... — reactive:us-ai-policy-regulation
  22. [22] OpenAI Reveals GPT-5.6 Sol Cybersecurity Model, Restricts Early Access - Infosecurity Magazine — reactive:us-ai-policy-regulation
  23. [23] When AI Safety Becomes AI Feudalism — reactive:us-ai-policy-regulation (2026-06-29)
  24. [24] If the U.S. turns frontier AI into a domestic-only or government-approved club, it risks doing to AI what it would never... — reactive:europe-ai-sovereignty-deficit (2026-06-27)
  25. [25] China Wins: Why Open-Source AI Becomes the Default — and Why $TAO Matters — reactive:us-ai-policy-regulation (2026-06-28)
  26. [26] Dario Amodei just published a super long blog, calling for an urgent policy overhaul because he thinks frontier AI is mo… — Rohan Paul Twitter (2026-06-10)
  27. [27] Dario Amodei argues in his new blog today, that powerful AI will soon become the dominant source of military and economi… — Rohan Paul Twitter (2026-06-11)
  28. [28] Anthropic is running a sophisticated regulatory capture strategy based on fear-mongering. It is principally responsible for the state regulatory frenzy that is damaging the startup ecosystem. — reactive:us-ai-policy-regulation
  29. [29] US VP JD Vance on AI Company's marketing strategy, some AI CEOs sell fear because fear makes their product look powerful… — Rohan Paul Twitter (2026-06-19)
  30. [30] CNN: Peter Thiel directly criticized the Vatican, saying Pope Leo XIV was unintentionally acting as a “Chinese communist… — Rohan Paul Twitter (2026-07-03)
  31. [31] Illinois Signs Frontier AI Safety Bill into Law | Chris Lehane posted ... — reactive:us-ai-policy-regulation
  32. [32] CA Senate Approves No Robo Bosses Act of 2026 to Ensure Human Oversight of AI in the Workplace | Senator Jerry McNerney — reactive:us-ai-policy-regulation
  33. [33] Bill Text: CA SB951 | 2025-2026 | Regular Session | Amended — reactive:us-ai-policy-regulation
  34. [34] 😺 42 states just subpoenaed OpenAI — The Neuron (2026-06-15)
  35. [35] Florida AG Sues OpenAI, Alleges ChatGPT Aided Mass Shooters — reactive:us-ai-policy-regulation
  36. [36] President Trump Signs Executive Order Preempting State AI Laws ... — reactive:us-ai-policy-regulation
  37. [37] President Trump Signs Executive Order Challenging State AI Laws | Paul Hastings LLP — reactive:us-ai-policy-regulation
  38. [38] Trump AI Executive Order Gives NSA Classified Role Over Frontier Models — reactive:us-ai-policy-regulation
  39. [39] Bloomberg, “Anthropic's AI Principles Make It a White House Target” — reactive:us-ai-policy-regulation
  40. [40] OpenAI Limits GPT-5.6 Rollout at US Government's Request — reactive:us-ai-policy-regulation
  41. [41] David Sacks is done being polite about Anthropic (Save this). — Milk Road AI Twitter (2026-06-14)
  42. [42] Sacks urges focus on Mythos-driven security needs, seeks to ‘correct’ the narrative around his Anthropic critique | InsideAIPolicy.com — reactive:us-ai-policy-regulation
  43. [43] Anthropic CEO disputes David Sacks' claims that company is 'woke' — reactive:us-ai-policy-regulation
  44. [44] Anthropic CEO defends support for AI regulations, alignment with ... — reactive:us-ai-policy-regulation
  45. [45] Anthropic CEO claps back after Trump officials accuse firm of AI fear-mongering   | TechCrunch — reactive:us-ai-policy-regulation
  46. [46] Legal Considerations Related to the Anthropic “Export Controls ... — reactive:us-ai-policy-regulation
  47. [47] OpenAI is trying to turn the regulator into a shareholder. Once the state is financially long OpenAI, every model-releas... — reactive:us-ai-policy-regulation (2026-07-02)
  48. [48] OpenAI’s Frontier Governance Framework — OpenAI Blog (2026-05-28)
  49. [49] OpenAI public policy agenda — OpenAI Blog (2026-06-03)
  50. [50] A blueprint for democratic governance of frontier AI — OpenAI Blog (2026-06-03)
  51. [51] 🟡 Let’s get physical — Semafor Technology (2026-06-03)
  52. [52] AI #171: False Flag — Zvi's AI Roundups (2026-06-04)
  53. [53] The OpenAI-Andreessen-Palantir SuperPAC admits that it was “part ... — reactive:us-ai-policy-regulation
  54. [54] Build American AI admits hiring an outside vendor to run anonymous sockpuppet accounts targeting its opponents · Digg — reactive:us-ai-policy-regulation
  55. [55] OpenAI releases GPT-5.6 with government as gatekeeper - LinkedIn — reactive:us-ai-policy-regulation
  56. [56] GPT-5.6 Preview System Card - OpenAI Deployment Safety Hub — reactive:us-ai-policy-regulation
  57. [57] NYT: The Trump administration is pressuring Meta to submit its AI models to government review before public release. — Rohan Paul Twitter (2026-06-24)
  58. [58] Meta Faces White House Push to Join AI Pre-Release ... — reactive:us-ai-policy-regulation
  59. [59] US Government Pressures Meta to Allow Pre-Release Reviews of Its ... — reactive:us-ai-policy-regulation
  60. [60] Meta Faces Growing Pressure From Trump Administration To Submit Advanced AI Models For Federal Testing: R - Benzinga — reactive:us-ai-policy-regulation
  61. [61] Trump administration pushes Meta for AI model security reviews- NYT — reactive:us-ai-policy-regulation
  62. [62] The Trump administration is leaning on Meta to agree to voluntary government review of its AI models — and as of today, ... — reactive:us-ai-policy-regulation (2026-06-24)
  63. [63] Meta Under Pressure to Join Federal AI Model Review Program - Techstrong.ai — reactive:us-ai-policy-regulation
  64. [64] Trump Administration Presses Meta To Submit AI Models For Federal Review — reactive:us-ai-policy-regulation
  65. [65] US Government Reportedly Urging Meta To Share Its AI Models — reactive:us-ai-policy-regulation
  66. [66] Meta holds out on government security reviews of its AI models — reactive:us-ai-policy-regulation
  67. [67] US presses Meta on AI model review - LinkedIn — reactive:us-ai-policy-regulation
  68. [68] Meta resists government review of AI models - LinkedIn — reactive:us-ai-policy-regulation
  69. [69] US presses Meta to agree to AI reviews as security concerns rise, NYT reports — reactive:us-ai-policy-regulation
  70. [70] SB 951: Employment: technological displacement: notice. — reactive:us-ai-policy-regulation
  71. [71] California Set to Restrict AI Use in the Workplace With “No Robo ... — reactive:us-ai-policy-regulation
  72. [72] Trump loses more control over AI regulation as Illinois passes landmark law — Ars Technica AI (2026-05-28)
  73. [73] Illinois Legislature passes historic AI bill that would require third-party safety audits — reactive:us-ai-policy-regulation
  74. [74] ChatGPT creators knew product would cause harm, Florida argues in lawsuit • Florida Phoenix — reactive:us-ai-policy-regulation
  75. [75] Attorney General James Uthmeier Launches Criminal Investigation into OpenAI, ChatGPT | My Florida Legal — reactive:us-ai-policy-regulation
  76. [76] Florida AG brings 'first-in-the-nation state-led lawsuit' against OpenAI and its CEO — reactive:us-ai-policy-regulation
  77. [77] Florida Attorney General Sues OpenAI Over ChatGPT's Alleged ... — reactive:us-ai-policy-regulation
  78. [78] Florida attorney general sues OpenAI | Courthouse News Service — reactive:us-ai-policy-regulation
  79. [79] Florida sues OpenAI over ChatGPT risks Attorney General ... — reactive:us-ai-policy-regulation
  80. [80] AI and Liability — Simon Willison (2026-06-25)
  81. [81] WATCH: Trump explains why he postponed signing AI executive order — reactive:us-ai-policy-regulation
  82. [82] A Pro-AI Super PAC’s Secret Meme Sockpuppets — reactive:us-ai-policy-regulation
  83. [83] A Dark-Money Campaign Is Paying Influencers to Frame Chinese AI as a Threat | WIRED — reactive:us-ai-policy-regulation
  84. [84] [PDF] GPT-5.6 Preview System Card - Deployment Safety Hub — reactive:us-ai-policy-regulation