artificial-intelligence.giGlobal AI Monitor

United States AI Regulation & Governance

US12 active modules0 quiet

Last updated · 12 modules · 17 instruments · 52 claims · 80 sources in the cumulative register

Framework status
Executive Order 14179, Removing Barriers to American Leadership in Artificial Intelligence (Valid from 2025-01-23).
Main regulator

RegulatorNo statutory federal AI regulator. Executive policy hub: AI.gov (AI Action Plan); technical body: NIST

Next milestone
No upcoming dated milestone recorded
Record status
Live · Periodic active record · 80 sources (75 retrieved) · 20 under review · Not yet expert-reviewed
AI-assembled from cited sources
Evidence-gated · How we verify ·
DisclosureNot yet expert-reviewed · Not legal, compliance or investment advice. Facts are machine-assembled from cited public sources. Individual claims have not each been independently verified by a human reviewer. See methodology.

Source tiers: T1 enacted statute, regulation, or binding supervisory order · T2 official regulator guidance, consultation, or registry · T3 international standard-setter or peer-reviewed analysis · T4 practitioner commentary or secondary reporting. Confidence: high = corroborated by retrieved T1/T2 text · medium = single official source or secondary corroboration · low = preliminary or unverified. Full definitions →

Research checked · Updated
Record history

Research as of · Record last changed · Published · Page built

Jump to module (12)

#

Executive Order 14179, Removing Barriers to American Leadership in Artificial Intelligence (Valid from 2025-01-23).

2 instruments (2 current) · 0 obligations · 8 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 281 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

Instruments (2)
  • Executive Order 14179, Removing Barriers to American Leadership in Artificial Intelligence Valid from 2025-01-23
    Binding
  • Executive Order, Ensuring a National Policy Framework for Artificial Intelligence Valid from 2025-12-11
    Binding
Claims and evidence (8)

  • Derived: Probable Colorado SB 26-189, signed 14 May 2026, repealed and replaced SB 24-205 before its original effective date, reframing the regime as a narrower automated-decision-making-technology notice-and-transparency statute effective 1 January 2027, enforced exclusively by the Colorado Attorney General with no private right of action.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • regulations.ai T3 · Retrieved · retrieved 4 Nov 2026Signed into law on May 14, 2026, SB26-189 repealed and replaced SB24-205 entirely, before the latter's original effective date.
    • Consumer Financial Services Law Monitor T3 · Retrieved · retrieved 4 Nov 2026SB 26‑189 is enforced exclusively by the Colorado Attorney General under the Colorado Consumer Protection Act. A violation of the developer or deployer requirements is deemed a deceptive trade practice, but private parties cannot bring suit
    • Finnegan T3 · Retrieved · retrieved 4 Nov 2026Pre-use notice. Before using covered ADMT to materially influence a consequential decision, provide consumers with a clear and conspicuous notice disclosing that ADMT will be used and explaining how to obtain additional information.
    • National Law Review T3 · Retrieved · retrieved 4 Nov 2026If covered ADMT results in an “adverse outcome,” which means a denial, termination, material reduction in benefits, or significantly worse pricing, the deployer must, within 30 days, provide a plain-language description of the decision and the ADMT’s role

  • Derived: Probable On 11 December 2025, the President signed an Executive Order directing the Attorney General to establish an AI Litigation Task Force to challenge state AI laws, the Secretary of Commerce to evaluate and potentially withhold BEAD funding from states with conflicting AI laws, and the FTC to issue a preemption policy statement within 90 days, while preserving state authority over child safety, AI infrastructure and state procurement.

    Confidence: Confirmed · As of 11 Dec 2025 · Evidenced
    Evidence and sources
    • The White House T1 · Retrieved · retrieved 4 Nov 2026The Order directs the Attorney General to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation.

  • Derived: Probable A 11 December 2025 executive order directs the Attorney General to establish an AI Litigation Task Force to challenge state AI laws as unconstitutional, preempted or unlawful, and conditions BEAD funding on a state's AI regulatory posture.

    Confidence: Confirmed · As of 11 Dec 2025 · Evidenced
    Evidence and sources
    • The White House T1 · Retrieved · retrieved 4 Nov 2026The Order directs the Attorney General to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation.
    • Gibson Dunn T3 · Retrieved · retrieved 4 Nov 2026The Order directs several federal government initiatives designed to blunt, preempt, or challenge state laws

  • Derived: Probable On 11 December 2025 the President signed Executive Order 14365 directing federal agencies to challenge and evaluate state AI laws for preemption.

    Confidence: Confirmed · As of 11 Dec 2025 · Evidenced
    Evidence and sources
    • The White House T1 · Retrieved · retrieved 4 Nov 2026The Order directs the Attorney General to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation.

  • A bill (GUARDRAILS Act) was introduced in the US House of Representatives to repeal EO 14365 and bar federal preemption of state AI regulation.

    Confidence: Probable (capped: best source is T3) · As of 1 Mar 2026 · Evidenced
    Evidence and sources
    • Holland & Knight T3 · Retrieved · retrieved 4 Nov 2026introduced the Guaranteeing and Upholding Americans' Right to Decide Responsible AI Laws and Standards (GUARDRAILS) Act

  • Derived: Probable EO 14179 (23 January 2025) revoked EO 14110, eliminating mandatory federal AI reporting requirements for higher-risk AI developers.

    Confidence: Probable (capped: best source is T3) · As of 31 Jan 2025 · Evidenced
    Evidence and sources
    • Hunton Andrews Kurth T3 · Retrieved · retrieved 4 Nov 2026By revoking the Biden Order, employers using AI will no longer be subject to such guidance or face mandatory federal reporting requirements.

  • A 11 December 2025 executive order directs federal agencies to challenge, defund and work toward preempting state AI laws judged onerous, while preserving state authority over child safety, AI infrastructure and state procurement.

    Confidence: Probable · As of 11 Dec 2025 · Evidenced
    Evidence and sources
    • The White House T1 · Retrieved · retrieved 4 Nov 2026The Order directs the Attorney General to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation.

  • The July 2025 America's AI Action Plan recommended withholding federal AI-related funding from states with burdensome AI regulations, a policy later implemented via the December 2025 executive order.

    Confidence: Probable (capped: best source is T3) · As of 12 Dec 2025 · Evidenced
    Evidence and sources
    • Skadden T3 · Retrieved · retrieved 2 Nov 2026the July 2025 policy blueprint “Winning the AI Race: America’s AI Action Plan” (AI Policy Blueprint), which called on the federal government not to direct AI-related funding to states with “burdensome” AI regulations.

#

TAKE IT DOWN Act (Valid from 2025-05-19).

2 instruments (2 current) · 0 obligations · 2 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 61 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

The federal TAKE IT DOWN Act, effective 19 May 2025, criminalises publishing or threatening to publish nonconsensual intimate imagery, including AI-generated deepfakes. In Texas, the Texas Responsible Artificial Intelligence Governance Act (TRAIGA, HB 149), effective 1 January 2026, prohibits developing or deploying AI with intent to unlawfully discriminate against a protected class (CLM-US-e2ea26d9, reported with Probable confidence).

Instruments (2)
  • TAKE IT DOWN Act Valid from 2025-05-19
    Binding
  • Texas Responsible Artificial Intelligence Governance Act (HB 149) Valid from 2026-01-01
    Binding
Claims and evidence (2)

  • Derived: Probable TRAIGA, effective 1 January 2026, prohibits developing or deploying AI with intent to unlawfully discriminate against a protected class.

    Confidence: Probable (capped: best source is T3) · As of 24 Jun 2025 · Evidenced
    Evidence and sources
    • K&L Gates T3 · Retrieved · retrieved 4 Nov 2026TRAIGA only prohibits the use of AI systems that are developed or deployed “with the intent to unlawfully discriminate against a protected class”

  • Derived: Probable The federal TAKE IT DOWN Act, effective 19 May 2025, criminalises publishing or threatening to publish nonconsensual intimate imagery, including AI deepfakes.

    Confidence: Probable · As of 20 May 2025 · Evidenced
    Evidence and sources
    • Congressional Research Service T1 · Surfaced · retrieved 4 Nov 2026the TAKE IT DOWN Act, a bill that criminalizes the nonconsensual publication of intimate images, including "digital forgeries" (i.e., deepfakes), in certain circumstances.

#

California CCPA Automated Decisionmaking Technology Regulations (CPPA, Article 11) (Valid from 2026-01-01).

2 instruments (2 current) · 4 obligations · 11 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 201 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

California's CCPA Automated Decisionmaking Technology Regulations (CPPA, Article 11) took effect 1 January 2026. Substantive ADMT compliance duties are phased in from 1 January 2027: businesses using ADMT for significant decisions must give consumers pre-use notice, offer at least two opt-out methods, and respond to access requests covering ADMT purpose, logic and the human role in the decision (§§ 7000-7302, § 7222). Separately, California's FEHA Automated-Decision System Regulations require employers, from 1 January 2027, to retain Covered ADMT inputs and outputs for four years (CLM-US-a350e4f0, reported with Probable confidence).

Colorado's original high-risk AI duty statute, SB 24-205, was repealed and replaced by SB 26-189 before SB 24-205's original effective date; SB 26-189 was signed 14 May 2026 (CLM-US-fd6be47a, CLM-US-af1cd61f, CLM-US-8f3081d6, reported with Probable confidence).

Instruments (2)
  • California CCPA Automated Decisionmaking Technology Regulations (CPPA, Article 11) Valid from 2026-01-01
    Binding
  • Colorado Automated Decision-Making Technology Act (SB 26-189) Valid from 2026-05-14
    Binding

Obligations

  • pre-use notice (deployer)

    Before using covered ADMT to materially influence a consequential decision, provide consumers with a clear and conspicuous notice disclosing that ADMT will be used.

    Transparency disclosure · Applies to: Deployer · Colorado Automated Decision-Making Technology Act (SB 26-189)
  • §§ 7000-7302; § 7222

    Pre-use notice, at least two opt-out methods, and access-request response on ADMT purpose/logic/human role for significant decisions.

    Transparency disclosure · Applies to: Deployer · California CCPA Automated Decisionmaking Technology Regulations (CPPA, Article 11)
  • developer documentation duty

    Provide deployers technical documentation on intended uses, training-data categories, limitations and human-review instructions.

    Documentation · Applies to: Product manufacturer · Colorado Automated Decision-Making Technology Act (SB 26-189)
  • deployer pre-use notice duty

    Clear and conspicuous pre-use notice before materially influencing a consequential decision with covered ADMT.

    Transparency disclosure · Applies to: Deployer · Colorado Automated Decision-Making Technology Act (SB 26-189)
Claims and evidence (11)

  • Derived: Probable Colorado SB 26-189, signed 14 May 2026, repealed and replaced SB 24-205 before its original effective date, reframing the regime as a narrower automated-decision-making-technology notice-and-transparency statute effective 1 January 2027, enforced exclusively by the Colorado Attorney General with no private right of action.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • regulations.ai T3 · Retrieved · retrieved 4 Nov 2026Signed into law on May 14, 2026, SB26-189 repealed and replaced SB24-205 entirely, before the latter's original effective date.
    • Consumer Financial Services Law Monitor T3 · Retrieved · retrieved 4 Nov 2026SB 26‑189 is enforced exclusively by the Colorado Attorney General under the Colorado Consumer Protection Act. A violation of the developer or deployer requirements is deemed a deceptive trade practice, but private parties cannot bring suit
    • Finnegan T3 · Retrieved · retrieved 4 Nov 2026Pre-use notice. Before using covered ADMT to materially influence a consequential decision, provide consumers with a clear and conspicuous notice disclosing that ADMT will be used and explaining how to obtain additional information.
    • National Law Review T3 · Retrieved · retrieved 4 Nov 2026If covered ADMT results in an “adverse outcome,” which means a denial, termination, material reduction in benefits, or significantly worse pricing, the deployer must, within 30 days, provide a plain-language description of the decision and the ADMT’s role

  • Derived: Probable Colorado's SB 24-205 high-risk AI duty statute was repealed and replaced by SB 26-189, signed 14 May 2026, before SB 24-205's original effective date.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • regulations.ai T3 · Retrieved · retrieved 4 Nov 2026Signed into law on May 14, 2026, SB26-189 repealed and replaced SB24-205 entirely, before the latter's original effective date.
    • Brownstein Hyatt Farber Schreck T3 · Retrieved · retrieved 4 Nov 2026SB 24-205 distinguishes between two categories of covered entities: Developers: entities that develop or substantially modify a high-risk artificial intelligence system.

  • Derived: Probable Colorado's SB 26-189 requires deployers of covered ADMT to give consumers clear and conspicuous pre-use notice before the technology is used to materially influence a consequential decision, effective 1 January 2027.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • Finnegan T3 · Retrieved · retrieved 4 Nov 2026Pre-use notice. Before using covered ADMT to materially influence a consequential decision, provide consumers with a clear and conspicuous notice disclosing that ADMT will be used and explaining how to obtain additional information.
    • Promise Legal T3 · Retrieved · retrieved 4 Nov 2026Deployers may rely on existing FCRA or FERPA consumer notices if those notices meet the ADMT Act's notice requirements.

  • Derived: Probable California's CCPA ADMT regulations require businesses using ADMT for significant decisions to provide pre-use notice, opt-out methods and access-request responses, with compliance required from 1 January 2027.

    Confidence: Confirmed · As of 23 Sep 2025 · Evidenced
    Evidence and sources
    • California Privacy Protection Agency T1 · Retrieved · retrieved 2 Nov 2026Businesses that use ADMT to make significant decisions must comply with the ADMT requirements beginning January 1, 2027.
    • California Privacy Protection Agency T1 · Retrieved · retrieved 4 Nov 2026“Automated decisionmaking technology” or “ADMT” means any technology that processes personal information and uses computation to replace human decisionmaking or substantially replace human decisionmaking.
    • Thompson Coburn T3 · Retrieved · retrieved 4 Nov 2026ADMT requirements apply; ADMT used for significant decisions prior to this date must be brought into compliance

  • Derived: Probable California's CPPA ADMT regulations took effect 1 January 2026, with substantive ADMT compliance obligations phased in from 1 January 2027.

    Confidence: Confirmed · As of 22 Sep 2025 · Evidenced
    Evidence and sources
    • California Privacy Protection Agency T1 · Retrieved · retrieved 4 Nov 2026implemented consumers' rights to access and opt–out of businesses' use of ADMT
    • Thompson Coburn T3 · Retrieved · retrieved 4 Nov 2026ADMT requirements apply; ADMT used for significant decisions prior to this date must be brought into compliance

  • Derived: Probable Colorado's SB 26-189, signed 14 May 2026, repealed and replaced the original Colorado AI Act (SB 24-205) before it took effect.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • regulations.ai T3 · Retrieved · retrieved 4 Nov 2026Signed into law on May 14, 2026, SB26-189 repealed and replaced SB24-205 entirely, before the latter's original effective date.

  • From 1 January 2027, Colorado deployers must give consumers clear pre-use notice before covered ADMT materially influences a consequential decision.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • Finnegan T3 · Retrieved · retrieved 4 Nov 2026Pre-use notice. Before using covered ADMT to materially influence a consequential decision, provide consumers with a clear and conspicuous notice disclosing that ADMT will be used and explaining how to obtain additional information.

  • From 1 January 2027, California businesses using ADMT for significant decisions must give pre-use notice, offer at least two opt-out methods, and answer access requests on ADMT purpose, logic and human role.

    Confidence: Confirmed · As of 23 Sep 2025 · Evidenced
    Evidence and sources
    • California Privacy Protection Agency T1 · Retrieved · retrieved 2 Nov 2026Businesses that use ADMT to make significant decisions must comply with the ADMT requirements beginning January 1, 2027.

  • From 1 January 2027, California FEHA regulations require employers to retain Covered ADMT inputs and outputs for four years.

    Confidence: Probable (capped: best source is T3) · As of 27 Mar 2026 · Evidenced
    Evidence and sources
    • Littler Mendelson T3 · Retrieved · retrieved 2 Nov 2026the new automated-decision system regulations under California’s Fair Employment and Housing Act will generally require retention of inputs to, and outputs from, Covered ADMT for four years.

  • From 1 January 2027, Colorado's ADMT Act requires developers of covered ADMT to give deployers technical documentation on intended use, training data, limitations and human-review instructions.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • Consumer Financial Services Law Monitor T3 · Retrieved · retrieved 4 Nov 2026If certain information is withheld because it is a trade secret or otherwise legally protected, the developer must notify the deployer.

  • From 1 January 2027, Colorado's ADMT Act requires deployers to give clear and conspicuous pre-use notice before using covered ADMT in a consequential decision.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • Finnegan T3 · Retrieved · retrieved 4 Nov 2026Pre-use notice. Before using covered ADMT to materially influence a consequential decision, provide consumers with a clear and conspicuous notice disclosing that ADMT will be used and explaining how to obtain additional information.

#

California Transparency in Frontier Artificial Intelligence Act (SB 53) (Valid from 2026-01-01).

1 instrument (1 current) · 1 obligation · 4 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 144 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

California's Transparency in Frontier Artificial Intelligence Act (SB 53, TFAIA) is in force from 1 January 2026. Frontier developers training models above the 10^26 FLOP compute threshold must publish a frontier AI framework and report critical safety incidents to California's Office of Emergency Services, with core publication, reporting and whistleblower duties effective from that date (CLM-US-88b80e3f, CLM-US-90eecd72, reported with Probable confidence though grounded partly in a T1 source). Separate OES and Attorney General reporting duties under SB 53 begin 1 January 2027 (CLM-US-8f476f1f, reported with Probable confidence).

California's Generative AI Training Data Transparency Act (AB 2013) requires developers of generative AI systems or services made available to Californians since 1 January 2022 to post a high-level training-data summary on their website, a duty in force from 1 January 2026 and subject to narrow security-integrity exemptions (CLM-US-5aa146c5, reported with Probable confidence).

Instruments (1)
  • California Transparency in Frontier Artificial Intelligence Act (SB 53) Valid from 2026-01-01
    Binding

Obligations

  • frontier framework publication and critical safety incident reporting

    Frontier developers must publish a frontier AI framework and report critical safety incidents to California's Office of Emergency Services.

    Incident reporting · Applies to: Other · California Transparency in Frontier Artificial Intelligence Act (SB 53)
Claims and evidence (4)

  • Under California AB 2013, developers of generative AI systems or services made available to Californians since 1 January 2022 must post a high-level training-data summary on their website from 1 January 2026, subject to narrow security-integrity exemptions.

    Confidence: Probable (capped: best source is T3) · As of 16 Jan 2026 · Evidenced
    Evidence and sources
    • Goodwin Procter T3 · Retrieved · retrieved 16 Jan 2026The TDTA requires AI developers to publicly post a “high-level summary” of the datasets used to train generative AI systems or services made available to the public since January 2022.
    • Jones Day T3 · Retrieved · retrieved 4 Nov 2026GenAI systems with the sole purpose of ensuring security and integrity, such as systems to detect security incidents.

  • Derived: Probable Under California SB 53 (TFAIA), frontier developers training models above the 10^26 FLOP compute threshold must publish a frontier AI framework and report critical safety incidents to the Office of Emergency Services, with core publication, reporting and whistleblower duties effective 1 January 2026.

    Confidence: Probable (capped: best source is T1) · As of 1 Oct 2025 · Evidenced
    Evidence and sources
    • K&L Gates T3 · Retrieved · retrieved 1 Oct 2025Core publication, reporting, truthfulness, and whistleblower obligations apply beginning January 1, 2026. OES and Attorney General anonymized reporting and annual reviews begin January 1, 2027.
    • LegiScan (California Legislature bill text) T1 · Surfaced · retrieved 4 Nov 2026The TFAIA would exempt from the California Public Records Act a report of a critical safety incident submitted to the Office of Emergency Services

  • Derived: Probable SB 53 requires frontier AI developers above a 10^26 FLOP compute threshold to publish a frontier AI framework and meet transparency, reporting and whistleblower duties, with core obligations effective 1 January 2026 and OES/AG reporting beginning 1 January 2027.

    Confidence: Probable (capped: best source is T3) · As of 1 Oct 2025 · Evidenced
    Evidence and sources
    • K&L Gates T3 · Retrieved · retrieved 1 Oct 2025Core publication, reporting, truthfulness, and whistleblower obligations apply beginning January 1, 2026. OES and Attorney General anonymized reporting and annual reviews begin January 1, 2027.
    • CASRAI T3 · Retrieved · retrieved 4 Nov 2026SB 53 defines "frontier developer" purely by training-compute threshold — more than 10^26 operations
    • Morrison Foerster T3 · Retrieved · retrieved 4 Nov 2026companies developing foundation models now face sweeping new compliance obligations—and penalties of up to $1 million per violation—effective January 1, 2026.

  • Derived: Probable California's TFAIA (SB 53) requires frontier AI developers to publish a safety framework and report critical safety incidents, with core duties effective 1 January 2026.

    Confidence: Probable (capped: best source is T3) · As of 1 Oct 2025 · Evidenced
    Evidence and sources
    • K&L Gates T3 · Retrieved · retrieved 1 Oct 2025Core publication, reporting, truthfulness, and whistleblower obligations apply beginning January 1, 2026. OES and Attorney General anonymized reporting and annual reviews begin January 1, 2027.

#

California AI Transparency Act (SB 942, as amended by AB 853) (Valid from 2026-08-02).

5 instruments (5 current) · 3 obligations · 12 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 355 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

The federal TAKE IT DOWN Act, effective 19 May 2025, requires covered online platforms to establish a mechanism for individuals to notify the platform of a nonconsensual intimate visual depiction, including AI-generated deepfakes, and to remove it, making reasonable efforts to remove known copies within 48 hours of a valid, verified request; failure to comply is treated as an FTC Act violation. The Federal Trade Commission began enforcing Section 3 of the Act on 19 May 2026 against platforms that fail to provide a compliant notice-and-removal mechanism (CLM-US-9396b902, CLM-US-1145887d, both Confirmed).

Utah's Artificial Intelligence Policy Act (UAIPA), in force since 1 May 2024, requires a person providing regulated-occupation services using generative AI to prominently disclose that use at the outset of the interaction (CLM-US-7c86c6c4, reported with Probable confidence); a subsequent amendment, SB 226, narrowed the mandatory disclosure trigger to 'high-risk' generative AI interactions only, leaving disclosure on request for other interactions (CLM-US-2e70084b, reported with Probable confidence).

Instruments (5)
  • California AI Transparency Act (SB 942, as amended by AB 853) Valid from 2026-08-02
    Binding
  • California Generative AI Training Data Transparency Act (AB 2013) Valid from 2026-01-01
    Binding
  • Colorado Automated Decision-Making Technology Act (SB 26-189) Valid from 2026-05-14
    Binding
  • TAKE IT DOWN Act Valid from 2025-05-19
    Binding
  • Utah Artificial Intelligence Policy Act (UAIPA) Valid from 2024-05-01
    Binding

Obligations

  • deployer pre-use notice duty

    Clear and conspicuous pre-use notice before materially influencing a consequential decision with covered ADMT.

    Transparency disclosure · Applies to: Deployer · Colorado Automated Decision-Making Technology Act (SB 26-189)
  • operative date and scope

    Content-provenance disclosure tools required for large generative AI providers from 2 August 2026.

    Marking or watermarking · Applies to: Product manufacturer · California AI Transparency Act (SB 942, as amended by AB 853)
  • Section 3 (platform notice-and-removal duty)

    Remove nonconsensual intimate depictions, including AI deepfakes, within 48 hours of a valid verified request.

    Other · Applies to: Other · TAKE IT DOWN Act
Claims and evidence (12)

  • Derived: Probable Colorado's SB 26-189 requires deployers of covered ADMT to give consumers clear and conspicuous pre-use notice before the technology is used to materially influence a consequential decision, effective 1 January 2027.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • Finnegan T3 · Retrieved · retrieved 4 Nov 2026Pre-use notice. Before using covered ADMT to materially influence a consequential decision, provide consumers with a clear and conspicuous notice disclosing that ADMT will be used and explaining how to obtain additional information.
    • Promise Legal T3 · Retrieved · retrieved 4 Nov 2026Deployers may rely on existing FCRA or FERPA consumer notices if those notices meet the ADMT Act's notice requirements.

  • Derived: Probable California's CCPA ADMT regulations require businesses using ADMT for significant decisions to provide pre-use notice, opt-out methods and access-request responses, with compliance required from 1 January 2027.

    Confidence: Confirmed · As of 23 Sep 2025 · Evidenced
    Evidence and sources
    • California Privacy Protection Agency T1 · Retrieved · retrieved 2 Nov 2026Businesses that use ADMT to make significant decisions must comply with the ADMT requirements beginning January 1, 2027.
    • California Privacy Protection Agency T1 · Retrieved · retrieved 4 Nov 2026“Automated decisionmaking technology” or “ADMT” means any technology that processes personal information and uses computation to replace human decisionmaking or substantially replace human decisionmaking.
    • Thompson Coburn T3 · Retrieved · retrieved 4 Nov 2026ADMT requirements apply; ADMT used for significant decisions prior to this date must be brought into compliance

  • Derived: Probable The FTC began enforcing Section 3 of the TAKE IT DOWN Act on 19 May 2026 against online platforms that fail to provide a notice-and-removal mechanism for nonconsensual intimate depictions, including AI-generated deepfakes.

    Confidence: Confirmed · As of 1 May 2026 · Evidenced
    Evidence and sources
    • Federal Trade Commission T1 · Retrieved · retrieved 2 Nov 2026the Federal Trade Commission begins enforcing Section 3 of the Act against platforms that fail to comply with the law's notice and removal requirements
    • ZwillGen T3 · Retrieved · retrieved 4 Nov 2026The law provides a limitation on liability against claims arising from the removal of apparent nonconsensual intimate visual depictions for platforms that implement and comply with this process

  • AB 2013 requires generative-AI developers serving California to publish a 12-category training-data summary from 1 January 2026, covering systems released or modified since 1 January 2022.

    Confidence: Probable (capped: best source is T3) · As of 16 Jan 2026 · Evidenced
    Evidence and sources
    • Goodwin Procter T3 · Retrieved · retrieved 16 Jan 2026The TDTA requires AI developers to publicly post a “high-level summary” of the datasets used to train generative AI systems or services made available to the public since January 2022.
    • Jones Day T3 · Retrieved · retrieved 4 Nov 2026GenAI systems with the sole purpose of ensuring security and integrity, such as systems to detect security incidents.

  • Derived: Probable The federal TAKE IT DOWN Act, effective 19 May 2025, criminalises publishing or threatening to publish nonconsensual intimate imagery, including AI deepfakes.

    Confidence: Probable · As of 20 May 2025 · Evidenced
    Evidence and sources
    • Congressional Research Service T1 · Surfaced · retrieved 4 Nov 2026the TAKE IT DOWN Act, a bill that criminalizes the nonconsensual publication of intimate images, including "digital forgeries" (i.e., deepfakes), in certain circumstances.

  • Derived: Probable California's AB 2013 requires generative AI developers to publicly post a training-data summary, effective 1 January 2026.

    Confidence: Probable (capped: best source is T3) · As of 16 Jan 2026 · Evidenced
    Evidence and sources
    • Goodwin Procter T3 · Retrieved · retrieved 16 Jan 2026The TDTA requires AI developers to publicly post a “high-level summary” of the datasets used to train generative AI systems or services made available to the public since January 2022.

  • NYC Local Law 144 requires employers to publish bias-audit summaries for six months and give candidates 10 business days' advance notice of AEDT use.

    Confidence: Probable (capped: best source is T3) · As of 28 Aug 2025 · Evidenced
    Evidence and sources
    • Pivot Point Security T3 · Retrieved · retrieved 2 Nov 2026Employers must publicly post audit summaries with start and audit dates, data sources, applicant counts, and key audit metrics for at least six months.

  • From 1 January 2027, Colorado's ADMT Act requires deployers to give clear and conspicuous pre-use notice before using covered ADMT in a consequential decision.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • Finnegan T3 · Retrieved · retrieved 4 Nov 2026Pre-use notice. Before using covered ADMT to materially influence a consequential decision, provide consumers with a clear and conspicuous notice disclosing that ADMT will be used and explaining how to obtain additional information.

  • Utah's AI Policy Act requires a person providing regulated-occupation services using generative AI to prominently disclose AI use at the outset of the interaction.

    Confidence: Probable (capped: best source is T3) · As of 1 May 2024 · Evidenced
    Evidence and sources
    • Hunton Andrews Kurth T3 · Retrieved · retrieved 2 Nov 2026Companies in violation of the AI Policy Act are subject to fines of up to $2,500 per violation and the Division of Consumer Protection may bring an action in court to enforce a provision of the law

  • Utah's SB 226 amendment narrowed UAIPA's mandatory disclosure trigger to 'high-risk' generative AI interactions only, with disclosure on request for other interactions.

    Confidence: Probable (capped: best source is T3) · As of 7 May 2025 · Evidenced
    Evidence and sources
    • Davis Polk T3 · Retrieved · retrieved 2 Nov 2026the UAIPA requires prominent disclosure at the outset of an interaction only if the use of Gen AI constitutes a “high-risk” AI interaction.

  • California's AI Transparency Act (SB 942, as amended by AB 853) requires content-provenance disclosure tools for large generative AI providers from 2 August 2026, its operative date having been moved from 1 January 2026.

    Confidence: Probable (capped: best source is T3) · As of 18 Aug 2026 · Evidenced
    Evidence and sources
    • Secure Privacy T3 · Retrieved · retrieved 2 Nov 2026AB 853, signed October 13, 2025, amended it and pushed the operative date to August 2, 2026, deliberately aligning with the EU AI Act's enforcement date for high-risk systems.

  • The TAKE IT DOWN Act requires covered platforms to remove AI-generated and other nonconsensual intimate depictions within 48 hours of a valid request, with FTC enforcement of this duty beginning 19 May 2026.

    Confidence: Confirmed · As of 1 May 2026 · Evidenced
    Evidence and sources
    • Federal Trade Commission T1 · Retrieved · retrieved 2 Nov 2026the Federal Trade Commission begins enforcing Section 3 of the Act against platforms that fail to comply with the law’s notice and removal requirements

#

California Generative AI Training Data Transparency Act (AB 2013) (Valid from 2026-01-01).

1 instrument (1 current) · 1 obligation · 3 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 66 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

California's Generative AI Training Data Transparency Act (AB 2013) requires developers of generative AI systems or services made available to Californians since 1 January 2022 to publicly post a high-level training-data summary, a duty in force from 1 January 2026 (CLM-US-31ef856e, CLM-US-5aa146c5, CLM-US-8fe18cbf, reported with Probable confidence), implemented at Cal. Bus. & Prof. Code § 3111 and covering a 12-category disclosure format.

Instruments (1)
  • California Generative AI Training Data Transparency Act (AB 2013) Valid from 2026-01-01
    Binding

Obligations

  • Cal. Bus. & Prof. Code § 3111

    Publicly post a high-level summary of training datasets for generative AI systems made available since 1 January 2022.

    Transparency disclosure · Applies to: Other · California Generative AI Training Data Transparency Act (AB 2013) · Deadline
Claims and evidence (3)

  • Under California AB 2013, developers of generative AI systems or services made available to Californians since 1 January 2022 must post a high-level training-data summary on their website from 1 January 2026, subject to narrow security-integrity exemptions.

    Confidence: Probable (capped: best source is T3) · As of 16 Jan 2026 · Evidenced
    Evidence and sources
    • Goodwin Procter T3 · Retrieved · retrieved 16 Jan 2026The TDTA requires AI developers to publicly post a “high-level summary” of the datasets used to train generative AI systems or services made available to the public since January 2022.
    • Jones Day T3 · Retrieved · retrieved 4 Nov 2026GenAI systems with the sole purpose of ensuring security and integrity, such as systems to detect security incidents.

  • AB 2013 requires generative-AI developers serving California to publish a 12-category training-data summary from 1 January 2026, covering systems released or modified since 1 January 2022.

    Confidence: Probable (capped: best source is T3) · As of 16 Jan 2026 · Evidenced
    Evidence and sources
    • Goodwin Procter T3 · Retrieved · retrieved 16 Jan 2026The TDTA requires AI developers to publicly post a “high-level summary” of the datasets used to train generative AI systems or services made available to the public since January 2022.
    • Jones Day T3 · Retrieved · retrieved 4 Nov 2026GenAI systems with the sole purpose of ensuring security and integrity, such as systems to detect security incidents.

  • Derived: Probable California's AB 2013 requires generative AI developers to publicly post a training-data summary, effective 1 January 2026.

    Confidence: Probable (capped: best source is T3) · As of 16 Jan 2026 · Evidenced
    Evidence and sources
    • Goodwin Procter T3 · Retrieved · retrieved 16 Jan 2026The TDTA requires AI developers to publicly post a “high-level summary” of the datasets used to train generative AI systems or services made available to the public since January 2022.

#

California CCPA Automated Decisionmaking Technology Regulations (CPPA, Article 11) (Valid from 2026-01-01).

8 instruments (8 current) · 6 obligations · 13 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 440 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

NYC Local Law 144 (Automated Employment Decision Tools), in force since 5 July 2023, bars use of an AEDT in employment screening or promotion decisions unless an independent bias audit has been conducted within the prior year (CLM-US-6669db4a, reported with Probable confidence), requires that auditor to be independent of the employer, the employment agency and the AEDT vendor (CLM-US-8571dfd4, Confirmed), and requires employers to publish bias-audit summaries for six months and give candidates 10 business days' advance notice of AEDT use (CLM-US-245b05f0, reported with Probable confidence).

Utah's Artificial Intelligence Policy Act provides that a company cannot escape liability for a Division of Consumer Protection statutory violation by attributing the violative statement or act to generative AI (CLM-US-02d44550, reported with Probable confidence).

Instruments (8)
  • California CCPA Automated Decisionmaking Technology Regulations (CPPA, Article 11) Valid from 2026-01-01
    Binding
  • California FEHA Automated-Decision System Regulations Date not recorded
    Binding
  • EEOC Title VII/ADA AI Technical Assistance (withdrawn) and Executive Order 14281 Valid from 2025-01-01
    Non-binding
  • Illinois Human Rights Act AI Amendment (HB 3773 / Public Act 103-0804) Valid from 2026-01-01
    Binding
  • NY DFS Insurance Circular Letter No. 7 (2024) Valid from 2024-07-11
    Regulatory expectation
  • NYC Local Law 144 (Automated Employment Decision Tools) Valid from 2023-07-05
    Binding
  • OMB Memorandum M-24-10 Valid from 2024-03-28
    Binding
  • Utah Artificial Intelligence Policy Act (UAIPA) Valid from 2024-05-01
    Binding

Obligations

  • accountability for generative AI violations

    A company using generative AI remains liable for a Division of Consumer Protection statute violation even if the AI made the violative statement or act.

    Other · Applies to: Other · Utah Artificial Intelligence Policy Act (UAIPA)
  • 20-871(a)(1); 6 RCNY 5-301

    Independent bias audit required before AEDT use in employment screening or promotion, repeated annually.

    Documentation · Applies to: Deployer · NYC Local Law 144 (Automated Employment Decision Tools)
  • publication and notice duties

    Publish bias-audit summary for six months and give 10 business days' advance notice before AEDT use.

    Transparency disclosure · Applies to: Deployer · NYC Local Law 144 (Automated Employment Decision Tools)
  • auditor independence; vendor liability allocation

    The AEDT vendor cannot perform the bias audit; auditor must be independent of employer, agency and vendor.

    Other · Applies to: Deployer · NYC Local Law 144 (Automated Employment Decision Tools)
  • 775 ILCS 5/2-102 (as amended)

    Discriminatory AI use, zip-code proxying, and failure to notify of AI use in covered employment decisions are civil rights violations.

    Prohibition · Applies to: Deployer · Illinois Human Rights Act AI Amendment (HB 3773 / Public Act 103-0804)
  • record retention requirement

    Retain Covered ADMT inputs/outputs for four years.

    Record keeping · Applies to: Deployer · California FEHA Automated-Decision System Regulations
Claims and evidence (13)

  • Derived: Probable Illinois HB 3773 made discriminatory or undisclosed AI use in employment a civil rights violation under the Illinois Human Rights Act effective 1 January 2026, notwithstanding IDHR's withdrawal of proposed implementing notice rules.

    Confidence: Probable (capped: best source is T3) · As of 20 Oct 2026 · Evidenced
    Evidence and sources
    • Secure Privacy T3 · Retrieved · retrieved 2 Nov 2026HB 3773 has made discriminatory or undisclosed AI use in employment a civil rights violation since January 1, 2026, even though the Illinois Department of Human Rights withdrew its draft notice rules in June.
    • Ogletree Deakins T3 · Retrieved · retrieved 4 Nov 2026HB 3773 amends the Illinois Human Rights Act, which applies to any person (with certain limited exceptions) employing one or more employees within Illinois during 20 or more calendar weeks
    • CASRAI T3 · Retrieved · retrieved 2 Nov 2026IDHR proposed the rules defining compliant notice on 15 May 2026 and withdrew them on 2 June 2026, with no refiling timeline.
    • Seyfarth Shaw T3 · Retrieved · retrieved 4 Nov 2026the underlying statutory obligations under HB 3773 remain in effect, while the contours of compliant notice in Illinois remain unclear.

  • Derived: Probable US export controls require a BIS licence for export of advanced AI computing hardware and certain AI model weights to restricted destinations.

    Confidence: Probable (capped: best source is T3) · As of 20 Oct 2026 · Evidenced
    Evidence and sources
    • Consumer Electronics Daily T3 · Retrieved · retrieved 4 Nov 2026BIS invoked GP10 formally in May 2025 guidance targeting Huawei Ascend AI chips, declaring that any use, export, reexport, or servicing of the Ascend 910B, 910C, or 910D constitutes an ongoing EAR violation absent prior BIS authorization.

  • Under Utah's AI Policy Act, a company cannot escape liability for a statutory violation by attributing the violative act or statement to generative AI.

    Confidence: Probable (capped: best source is T3) · As of 1 May 2024 · Evidenced
    Evidence and sources
    • Hunton Andrews Kurth T3 · Retrieved · retrieved 2 Nov 2026Companies in violation of the AI Policy Act are subject to fines of up to $2,500 per violation and the Division of Consumer Protection may bring an action in court to enforce a provision of the law

  • NYC Local Law 144 bars use of an AEDT in employment screening or promotion decisions unless an independent bias audit has been conducted within the prior year.

    Confidence: Probable (capped: best source is T3) · As of 5 Jul 2023 · Evidenced
    Evidence and sources
    • NHIMG T3 · Retrieved · retrieved 2 Nov 2026New York City's Local Law 144 makes it unlawful to use an automated employment decision tool to screen candidates or employees unless the tool has had an independent bias audit in the last

  • NYC Local Law 144 requires employers to publish bias-audit summaries for six months and give candidates 10 business days' advance notice of AEDT use.

    Confidence: Probable (capped: best source is T3) · As of 28 Aug 2025 · Evidenced
    Evidence and sources
    • Pivot Point Security T3 · Retrieved · retrieved 2 Nov 2026Employers must publicly post audit summaries with start and audit dates, data sources, applicant counts, and key audit metrics for at least six months.

  • NYC Local Law 144 requires an independent auditor, free of financial or employment ties to the employer, agency or AEDT vendor, to perform the bias audit.

    Confidence: Confirmed · As of 18 Aug 2026 · Evidenced
    Evidence and sources

  • Since 1 January 2026, discriminatory or undisclosed AI use in covered Illinois employment decisions is a civil rights violation under the Illinois Human Rights Act as amended by HB 3773.

    Confidence: Probable (capped: best source is T3) · As of 20 Oct 2026 · Evidenced
    Evidence and sources
    • Secure Privacy T3 · Retrieved · retrieved 2 Nov 2026HB 3773 has made discriminatory or undisclosed AI use in employment a civil rights violation since January 1, 2026, even though the Illinois Department of Human Rights withdrew its draft notice rules in June.

  • Illinois withdrew its proposed Subpart J rules defining AI-use notice on 2 June 2026, leaving the underlying HB 3773 notice duty in force but without implementing-rule detail on compliant notice.

    Confidence: Probable (capped: best source is T3) · As of 20 Oct 2026 · Evidenced
    Evidence and sources
    • CASRAI T3 · Retrieved · retrieved 2 Nov 2026IDHR proposed the rules defining compliant notice on 15 May 2026 and withdrew them on 2 June 2026, with no refiling timeline.

  • From 1 January 2027, California businesses using ADMT for significant decisions must give pre-use notice, offer at least two opt-out methods, and answer access requests on ADMT purpose, logic and human role.

    Confidence: Confirmed · As of 23 Sep 2025 · Evidenced
    Evidence and sources
    • California Privacy Protection Agency T1 · Retrieved · retrieved 2 Nov 2026Businesses that use ADMT to make significant decisions must comply with the ADMT requirements beginning January 1, 2027.

  • From 1 January 2027, California FEHA regulations require employers to retain Covered ADMT inputs and outputs for four years.

    Confidence: Probable (capped: best source is T3) · As of 27 Mar 2026 · Evidenced
    Evidence and sources
    • Littler Mendelson T3 · Retrieved · retrieved 2 Nov 2026the new automated-decision system regulations under California’s Fair Employment and Housing Act will generally require retention of inputs to, and outputs from, Covered ADMT for four years.

  • Although the EEOC withdrew its AI technical-assistance guidance and has deprioritized disparate-impact enforcement under EO 14281, Title VII's disparate-impact prohibition remains codified and private plaintiffs retain standing to sue employers over AI-driven disparate impact.

    Confidence: Probable (capped: best source is T3) · As of 25 Jun 2025 · Evidenced
    Evidence and sources
    • National Law Review T3 · Retrieved · retrieved 2 Nov 2026Title VII's disparate impact prohibition and UGESP's validation requirements apply to AI-driven selection procedures regardless of whether the EEOC publishes guidance explaining how.

  • OMB Memorandum M-24-10 requires federal civilian agencies to designate a Chief AI Officer, govern rights- and safety-impacting AI, and publish public AI use-case inventories.

    Confidence: Probable (capped: best source is T3) · As of 28 Mar 2024 · Evidenced
    Evidence and sources
    • DeepInspect T3 · Retrieved · retrieved 2 Nov 2026OMB Memorandum M-24-10, issued March 28, 2024, set the AI governance baseline for federal civilian agencies including risk management for rights-impacting and safety-impacting AI, a Chief AI Officer designation, and public inventories of AI use cases.

  • NY DFS Circular Letter No. 7 sets supervisory expectations for insurers' use of AI and external consumer data in underwriting and pricing, including proxy-discrimination testing and disclosure duties.

    Confidence: Probable (capped: best source is T3) · As of 12 Jul 2024 · Evidenced
    Evidence and sources
    • Willkie Compliance Concourse T3 · Retrieved · retrieved 2 Nov 2026The Circular Letter is intended to provide DFS’s expectations for insurers authorized to write insurance in New York who develop and/or use external consumer data and information sources

#

California Transparency in Frontier Artificial Intelligence Act (SB 53) (Valid from 2026-01-01).

1 instrument (1 current) · 1 obligation · 3 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 94 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

California's Transparency in Frontier Artificial Intelligence Act (SB 53) requires frontier AI developers training models above the 10^26 FLOP compute threshold to publish a frontier AI framework and to report critical safety incidents to California's Office of Emergency Services, with core publication, reporting and whistleblower duties effective from 1 January 2026 (CLM-US-88b80e3f, CLM-US-90eecd72, reported with Probable confidence, grounded in part in a T1 source).

Enacted, not yet in force

Office of Emergency Services and Attorney General reporting duties under SB 53 begin 1 January 2027 (CLM-US-8f476f1f, reported with Probable confidence).

Instruments (1)
  • California Transparency in Frontier Artificial Intelligence Act (SB 53) Valid from 2026-01-01
    Binding

Obligations

  • frontier framework publication and critical safety incident reporting

    Frontier developers must publish a frontier AI framework and report critical safety incidents to California's Office of Emergency Services.

    Incident reporting · Applies to: Other · California Transparency in Frontier Artificial Intelligence Act (SB 53)
Claims and evidence (3)

  • Derived: Probable Under California SB 53 (TFAIA), frontier developers training models above the 10^26 FLOP compute threshold must publish a frontier AI framework and report critical safety incidents to the Office of Emergency Services, with core publication, reporting and whistleblower duties effective 1 January 2026.

    Confidence: Probable (capped: best source is T1) · As of 1 Oct 2025 · Evidenced
    Evidence and sources
    • K&L Gates T3 · Retrieved · retrieved 1 Oct 2025Core publication, reporting, truthfulness, and whistleblower obligations apply beginning January 1, 2026. OES and Attorney General anonymized reporting and annual reviews begin January 1, 2027.
    • LegiScan (California Legislature bill text) T1 · Surfaced · retrieved 4 Nov 2026The TFAIA would exempt from the California Public Records Act a report of a critical safety incident submitted to the Office of Emergency Services

  • Derived: Probable SB 53 requires frontier AI developers above a 10^26 FLOP compute threshold to publish a frontier AI framework and meet transparency, reporting and whistleblower duties, with core obligations effective 1 January 2026 and OES/AG reporting beginning 1 January 2027.

    Confidence: Probable (capped: best source is T3) · As of 1 Oct 2025 · Evidenced
    Evidence and sources
    • K&L Gates T3 · Retrieved · retrieved 1 Oct 2025Core publication, reporting, truthfulness, and whistleblower obligations apply beginning January 1, 2026. OES and Attorney General anonymized reporting and annual reviews begin January 1, 2027.
    • CASRAI T3 · Retrieved · retrieved 4 Nov 2026SB 53 defines "frontier developer" purely by training-compute threshold — more than 10^26 operations
    • Morrison Foerster T3 · Retrieved · retrieved 4 Nov 2026companies developing foundation models now face sweeping new compliance obligations—and penalties of up to $1 million per violation—effective January 1, 2026.

  • Derived: Probable California's TFAIA (SB 53) requires frontier AI developers to publish a safety framework and report critical safety incidents, with core duties effective 1 January 2026.

    Confidence: Probable (capped: best source is T3) · As of 1 Oct 2025 · Evidenced
    Evidence and sources
    • K&L Gates T3 · Retrieved · retrieved 1 Oct 2025Core publication, reporting, truthfulness, and whistleblower obligations apply beginning January 1, 2026. OES and Attorney General anonymized reporting and annual reviews begin January 1, 2027.

#

Colorado Automated Decision-Making Technology Act (SB 26-189) (Valid from 2026-05-14).

2 instruments (2 current) · 1 obligation · 2 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 78 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Non-binding frameworks and standards

The NIST AI Risk Management Framework 1.0 (AI RMF), published 26 January 2023, remains voluntary, non-binding guidance, though it is referenced as a compliance baseline in some state AI safe-harbor provisions (CLM-US-4e8dd1cf, Confirmed).

Enacted, not yet in force

Colorado's SB 26-189 requires developers of covered automated decision-making technology to give deployers technical documentation on intended use, training data, limitations and human-review instructions, a duty taking effect 1 January 2027 (CLM-US-105153b6, reported with Probable confidence).

Instruments (2)
  • Colorado Automated Decision-Making Technology Act (SB 26-189) Valid from 2026-05-14
    Binding
  • NIST AI Risk Management Framework 1.0 (AI RMF) Valid from 2023-01-26
    Non-binding

Obligations

  • developer documentation duty

    Provide deployers technical documentation on intended uses, training-data categories, limitations and human-review instructions.

    Documentation · Applies to: Product manufacturer · Colorado Automated Decision-Making Technology Act (SB 26-189)
Claims and evidence (2)

  • The NIST AI Risk Management Framework is voluntary, non-binding guidance, though referenced as a compliance baseline in some state AI safe-harbor provisions.

    Confidence: Confirmed · As of 26 Jan 2023 · Evidenced
    Evidence and sources
    • National Institute of Standards and Technology T1 · Retrieved · retrieved 2 Nov 2026The NIST AI Risk Management Framework (AI RMF) is intended for voluntary use and to improve the ability to incorporate trustworthiness considerations into the design, development, use, and evaluation of AI products, services, and systems.

  • From 1 January 2027, Colorado's ADMT Act requires developers of covered ADMT to give deployers technical documentation on intended use, training data, limitations and human-review instructions.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • Consumer Financial Services Law Monitor T3 · Retrieved · retrieved 4 Nov 2026If certain information is withheld because it is a trade secret or otherwise legally protected, the developer must notify the deployer.

#

Colorado Automated Decision-Making Technology Act (SB 26-189) (Valid from 2026-05-14).

4 instruments (4 current) · 0 obligations · 11 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 346 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force and enforcement activity

The Federal Trade Commission ordered Rite Aid Corporation, which had deployed AI-based facial recognition surveillance that falsely flagged consumers and disproportionately affected women and people of colour, to be prohibited from using facial recognition technology for surveillance purposes for five years, by consent order dated 19 December 2023 (CLM-US-a415e59d, Confirmed); the FTC separately opened an investigation into Rite Aid's deployment of AI facial recognition surveillance without reasonable safeguards. The FTC brought at least a dozen deceptive AI-capability-claim ('AI-washing') enforcement actions in 2025 under FTC Act Section 5 (CLM-US-e2195e76, Confirmed). The FTC also reopened and set aside its prior consent order against Rytr, an AI tool that generated fake consumer reviews, on 22 December 2025, citing the administration's AI Action Plan.

Instruments (4)
  • Colorado Automated Decision-Making Technology Act (SB 26-189) Valid from 2026-05-14
    Binding
  • EEOC Title VII/ADA AI Technical Assistance (withdrawn) and Executive Order 14281 Valid from 2025-01-01
    Non-binding
  • Executive Order, Ensuring a National Policy Framework for Artificial Intelligence Valid from 2025-12-11
    Binding
  • TAKE IT DOWN Act Valid from 2025-05-19
    Binding
Claims and evidence (11)

  • Derived: Probable On 11 December 2025, the President signed an Executive Order directing the Attorney General to establish an AI Litigation Task Force to challenge state AI laws, the Secretary of Commerce to evaluate and potentially withhold BEAD funding from states with conflicting AI laws, and the FTC to issue a preemption policy statement within 90 days, while preserving state authority over child safety, AI infrastructure and state procurement.

    Confidence: Confirmed · As of 11 Dec 2025 · Evidenced
    Evidence and sources
    • The White House T1 · Retrieved · retrieved 4 Nov 2026The Order directs the Attorney General to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation.

  • Derived: Probable The FTC ordered Rite Aid Corporation, which deployed AI-based facial recognition surveillance that falsely flagged consumers disproportionately affecting women and people of color, to be prohibited from using facial recognition technology for surveillance purposes for five years.

    Confidence: Confirmed · As of 19 Dec 2023 · Evidenced
    Evidence and sources
    • Federal Trade Commission T1 · Retrieved · retrieved 4 Nov 2026Rite Aid will be prohibited from using facial recognition technology for surveillance purposes for five years

  • Derived: Probable A 11 December 2025 executive order directs the Attorney General to establish an AI Litigation Task Force to challenge state AI laws as unconstitutional, preempted or unlawful, and conditions BEAD funding on a state's AI regulatory posture.

    Confidence: Confirmed · As of 11 Dec 2025 · Evidenced
    Evidence and sources
    • The White House T1 · Retrieved · retrieved 4 Nov 2026The Order directs the Attorney General to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation.
    • Gibson Dunn T3 · Retrieved · retrieved 4 Nov 2026The Order directs several federal government initiatives designed to blunt, preempt, or challenge state laws

  • Derived: Probable The FTC began enforcing Section 3 of the TAKE IT DOWN Act on 19 May 2026 against online platforms that fail to provide a notice-and-removal mechanism for nonconsensual intimate depictions, including AI-generated deepfakes.

    Confidence: Confirmed · As of 1 May 2026 · Evidenced
    Evidence and sources
    • Federal Trade Commission T1 · Retrieved · retrieved 2 Nov 2026the Federal Trade Commission begins enforcing Section 3 of the Act against platforms that fail to comply with the law's notice and removal requirements
    • ZwillGen T3 · Retrieved · retrieved 4 Nov 2026The law provides a limitation on liability against claims arising from the removal of apparent nonconsensual intimate visual depictions for platforms that implement and comply with this process

  • Derived: Probable On 11 December 2025 the President signed Executive Order 14365 directing federal agencies to challenge and evaluate state AI laws for preemption.

    Confidence: Confirmed · As of 11 Dec 2025 · Evidenced
    Evidence and sources
    • The White House T1 · Retrieved · retrieved 4 Nov 2026The Order directs the Attorney General to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation.

  • SB 26-189 violations are deceptive trade practices enforced solely by the Colorado AG, with no private right of action and generally a 60-day cure period.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • Consumer Financial Services Law Monitor T3 · Retrieved · retrieved 4 Nov 2026SB 26‑189 is enforced exclusively by the Colorado Attorney General under the Colorado Consumer Protection Act. A violation of the developer or deployer requirements is deemed a deceptive trade practice, but private parties cannot bring suit

  • The FTC brought at least a dozen deceptive AI-capability-claim ('AI-washing') enforcement actions in 2025 under FTC Act Section 5.

    Confidence: Probable · As of 2 Apr 2026 · Evidenced
    Evidence and sources
    • Hunton Andrews Kurth T3 · Retrieved · retrieved 4 Nov 2026The agency brought at least a dozen AI-washing cases in 2025
    • Federal Trade Commission T1 · Retrieved · retrieved 4 Nov 2026FTC Reopens and Sets Aside Rytr Final Order in Response to the Trump Administration's AI Action Plan

  • A coalition of 35 US state Attorneys General sent a letter to AI companies on 25 September 2025 raising child-safety and deepfake concerns.

    Confidence: Probable (capped: best source is T3) · As of 25 Sep 2025 · Evidenced
    Evidence and sources
    • AI Lawsuit Tracker T3 · Retrieved · retrieved 4 Nov 2026A coalition of 35 state attorneys general sent a letter to leading AI companies

  • A 11 December 2025 executive order directs federal agencies to challenge, defund and work toward preempting state AI laws judged onerous, while preserving state authority over child safety, AI infrastructure and state procurement.

    Confidence: Probable · As of 11 Dec 2025 · Evidenced
    Evidence and sources
    • The White House T1 · Retrieved · retrieved 4 Nov 2026The Order directs the Attorney General to establish an AI Litigation Task Force to challenge unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation.

  • Although the EEOC withdrew its AI technical-assistance guidance and has deprioritized disparate-impact enforcement under EO 14281, Title VII's disparate-impact prohibition remains codified and private plaintiffs retain standing to sue employers over AI-driven disparate impact.

    Confidence: Probable (capped: best source is T3) · As of 25 Jun 2025 · Evidenced
    Evidence and sources
    • National Law Review T3 · Retrieved · retrieved 2 Nov 2026Title VII's disparate impact prohibition and UGESP's validation requirements apply to AI-driven selection procedures regardless of whether the EEOC publishes guidance explaining how.

  • The TAKE IT DOWN Act requires covered platforms to remove AI-generated and other nonconsensual intimate depictions within 48 hours of a valid request, with FTC enforcement of this duty beginning 19 May 2026.

    Confidence: Confirmed · As of 1 May 2026 · Evidenced
    Evidence and sources
    • Federal Trade Commission T1 · Retrieved · retrieved 2 Nov 2026the Federal Trade Commission begins enforcing Section 3 of the Act against platforms that fail to comply with the law’s notice and removal requirements

#

Texas Responsible Artificial Intelligence Governance Act (HB 149) (Valid from 2026-01-01).

1 instrument (1 current) · 0 obligations · 2 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 77 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

Binding law in force

The Texas Responsible Artificial Intelligence Governance Act (TRAIGA, HB 149), effective 1 January 2026, establishes a 36-month regulatory sandbox and a NIST AI Risk Management Framework-based safe harbour from enforcement (CLM-US-e6f344cb, reported with Probable confidence).

Colorado's original high-risk AI duty statute, SB 24-205, was repealed and replaced by SB 26-189, signed 14 May 2026, before SB 24-205's original effective date, providing transitional relief from the original regime's duties (CLM-US-fd6be47a, reported with Probable confidence).

Instruments (1)
  • Texas Responsible Artificial Intelligence Governance Act (HB 149) Valid from 2026-01-01
    Binding
Claims and evidence (2)

  • Derived: Probable Colorado's SB 24-205 high-risk AI duty statute was repealed and replaced by SB 26-189, signed 14 May 2026, before SB 24-205's original effective date.

    Confidence: Probable (capped: best source is T3) · As of 14 May 2026 · Evidenced
    Evidence and sources
    • regulations.ai T3 · Retrieved · retrieved 4 Nov 2026Signed into law on May 14, 2026, SB26-189 repealed and replaced SB24-205 entirely, before the latter's original effective date.
    • Brownstein Hyatt Farber Schreck T3 · Retrieved · retrieved 4 Nov 2026SB 24-205 distinguishes between two categories of covered entities: Developers: entities that develop or substantially modify a high-risk artificial intelligence system.

  • TRAIGA establishes a 36-month regulatory sandbox and a NIST AI RMF-based safe harbour from enforcement, effective 1 January 2026.

    Confidence: Probable (capped: best source is T3) · As of 7 Mar 2026 · Evidenced
    Evidence and sources
    • Modulos T3 · Retrieved · retrieved 4 Nov 2026TRAIGA establishes a 36-month regulatory sandbox administered by the Texas Department of Information Resources. Approved participants can test AI systems without standard state licensing requirements, enforcement actions, or punitive agency action.

#

On 28 May 2026, Anthropic closed a $65 billion Series H funding round at a $965 billion post-money valuation.

0 instruments (0 current) · 0 obligations · 4 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 80 words · last cycle 11 Nov 2026 · AI research, not expert reviewed

OpenAI closed a USD 122 billion funding round in March 2026 at a USD 852 billion valuation (CLM-US-d148ee71, reported with Probable confidence). Anthropic closed a USD 65 billion Series H funding round on 28 May 2026 at a USD 965 billion post-money valuation (CLM-US-8e832684, CLM-US-efa3ec79, reported with Probable confidence). OpenAI, Oracle and Related Digital held the groundbreaking for the USD 16 billion Stargate data-centre campus in Michigan on 1 June 2026 (CLM-US-82f68442, reported with Probable confidence at the T2 tier).

Claims and evidence (4)

  • On 28 May 2026, Anthropic closed a $65 billion Series H funding round at a $965 billion post-money valuation.

    Confidence: Probable (capped: best source is T3) · As of 28 May 2026 · Evidenced
    Evidence and sources
    • Pinggy.io T3 · Retrieved · retrieved 4 Nov 2026On May 28, 2026, Anthropic closed a $65 billion Series H at a $965 billion post-money valuation - briefly making it the most valuable private company in the world, surpassing OpenAI.

  • Anthropic closed a USD 65 billion Series H funding round on 28 May 2026 at a USD 965 billion post-money valuation.

    Confidence: Probable (capped: best source is T3) · As of 28 May 2026 · Evidenced
    Evidence and sources
    • Pinggy.io T3 · Retrieved · retrieved 4 Nov 2026On May 28, 2026, Anthropic closed a $65 billion Series H at a $965 billion post-money valuation - briefly making it the most valuable private company in the world, surpassing OpenAI.

  • OpenAI closed a USD 122 billion funding round in March 2026 at a USD 852 billion valuation.

    Confidence: Probable (capped: best source is T3) · As of 1 Mar 2026 · Evidenced
    Evidence and sources
    • Pinggy.io T3 · Retrieved · retrieved 4 Nov 2026On May 28, 2026, Anthropic closed a $65 billion Series H at a $965 billion post-money valuation - briefly making it the most valuable private company in the world, surpassing OpenAI.

  • OpenAI, Oracle and Related Digital held the groundbreaking for the USD 16 billion Stargate data-centre campus in Michigan on 1 June 2026.

    Confidence: Probable · As of 1 Jun 2026 · Evidenced
    Evidence and sources
    • CNBC T2 · Retrieved · retrieved 4 Nov 2026OpenAI and Oracle held an official groundbreaking for the $16 billion Stargate campus in Saline Township, Michigan, on Monday.
No modules match.

Filters combine as OR inside a group and AND across groups.

Authorities

No authorities recorded in this research.

Enforcement actions

DateRespondentAction
Rite Aid CorporationProbe
Rite Aid CorporationConsent order
RytrSettlement

Cases

No cases recorded in this research.

Consultations

No consultations recorded in this research.

What this record does not yet establish is set out on the transparency register; its dated events are on the regulatory calendar.

Think something on this page is wrong? Report an error.

Technical provenance (record schema, run ID & build metadata)
Schema version
artificial-intelligence-record-v2.12
Research run ID
aic-union-US-2026-11-04
Page model
module-card-page-v1
Renderer
render_jid_aic_v13
Build date