Does Trump AI Executive Order Cancel State AI Laws? Late-2026 Employer Scoreboard

A composite scenario assembled from documented cases: On December 15, 2025, an operations director at a 38-person Boston recruiting firm posted a news link in Slack with three words: "Hold the notices." Four days earlier, the White House had signed Executive Order 14365 directing federal agencies to challenge state AI statutes (Sidley, Dec 23, 2025). Assuming preemption was settled, the firm shelved its AI notice for Chicago applicants, kept an unvetted resume scorer running in Austin and Oakland, and renewed its January 2026 general liability policy unread. By February 2026, Illinois HB 3773 and Texas TRAIGA were both in force, California FEHA rules governed five-employee workplaces, and ISO endorsement CG 40 47 had stripped generative-AI liability coverage (IIABA, Oct 21, 2025).

Freezing compliance over preemption headlines exposes small employers to steep fines as insurers pull back coverage. Intentional uncurable discrimination under Texas TRAIGA carries $80,000 to $200,000 per violation (Latham, Jun 2025). Massachusetts secured a $2.5 million AI loan-underwriting settlement (Mass.gov, Jul 10, 2025), and a Boston class action over automated tenant scores settled for $2.275 million (Cohen Milstein). IBM's 2026 Cost of a Data Breach study (n = 602 breached organizations, vendor-sponsored Ponemon research) found unapproved staff AI tools in 43% of security incidents (IBM, Jul 29, 2026). Part of our safe use of AI tools at work hub and paired with our one-page AI use policy, this October 29, 2026 scoreboard is operational reference, not legal advice.

Why can't a presidential executive order unilaterally cancel state AI statutes?

Presidential executive orders, DOJ lawsuits, and proposed FTC policy statements cannot cancel enacted state AI statutes because the US Constitution's Supremacy Clause (Article VI, Clause 2) grants preemption power to federal laws passed by Congress and statutorily authorized regulations, not White House directives. Until Congress passes a preemption statute or a federal court enjoins a specific state law, state enforcers and private plaintiffs retain full statutory authority.

Think of municipal snow-emergency parking rules. Suppose a governor calls local towing ordinances burdensome, tells state lawyers to file test cases, and threatens to withhold highway grants. None of those statements repeals the town bylaw on your street; park in a snow zone tonight and the tow truck still hooks your bumper at 6:00 AM. Where that analogy breaks is multi-year administrative pressure. Executive Order 14365 (signed December 11, 2025) launched four federal tracks that challenge state rules over time without repealing any statute overnight:

1. DOJ AI Litigation Task Force: Sues states on Commerce Clause, First Amendment, Equal Protection, and preemption theories. On April 24, 2026, DOJ joined xAI's challenge against Colorado's AI law (Proskauer, Apr 28, 2026). Filing suit opens a docket, not an injunction.
2. FTC & FCC Proceedings: On July 1, 2026, the FTC voted 2 to 0 to seek comment on a proposed policy statement arguing state laws forcing ideological distortion of truthful AI outputs conflict with the FTC Act (FTC, Jul 1, 2026). Comments closed July 31, 2026; zero state HR or fraud statutes were voided.
3. Commerce & BEAD Pressure: Ordered Commerce to list "onerous" state AI laws by March 11, 2026 (S&P Global) and tie ~$21 billion in BEAD funds to deregulation (Mintz). Verify any list on .gov dockets rather than third-party wikis.
4. Stalled Congressional Bills: The Senate voted 99 to 1 in July 2025 to strip a 10-year state AI moratorium, and a June 4, 2026 House three-year development preemption draft remained stalled in October 2026 (CASRAI, Sep 20, 2026). EO 14365 itself carved out child safety, data-center rules, and state procurement.

Federal deception enforcement also continued: after setting aside its 2024 Rytr order, the FTC in 2026 hit Air AI with an $18 million judgment ($50,000 paid plus a permanent ban) over false human-replacement claims (CFO Dive), barred Growth Cave's AI revenue promises, and restricted Workado's unverified AI-detector claims.

Which US state AI employment and chatbot laws are in force as of October 29, 2026?

As of October 29, 2026, workplace and chatbot AI statutes are enforced in Illinois, Texas, California, and New York City, while overhauled notice, layoff, and human-corroboration mandates take effect in Colorado, California, and Connecticut through July 2027. Out of 48 workplace AI bills filed across 16 states in 2025-2026 (Tracker), five enacted regimes set the compliance floor:

• Illinois HB 3773 (IN FORCE Jan 1, 2026): Covers IHRA employers broadly (Hinshaw), barring AI disparate impact and ZIP proxies while making omitted AI notice an independent violation (draft Subpart J: tool, vendor, purpose, jobs, data, contact, ADA path; 4-year logs; IHRC fees).
• Texas TRAIGA (IN FORCE Jan 1, 2026): Covers firms operating in Texas or serving Texans (Latham), barring intentional AI discrimination (disparate impact alone insufficient). AG-only after 60-day cure; $80,000 to $200,000 uncurable fines; NIST AI RMF affirmative defense.
• California's Four-Part Stack:

1. FEHA ADS Rules (IN FORCE Oct 1, 2025): Covers 5+ staff, treats AI vendors as "agents" for disparate impact, and mandates 4-year selection logs (Mayer Brown).
2. SB 243 Companion Chatbots (IN FORCE Jan 1, 2026): Mandates AI disclosure, crisis protocols, and 3-hour minor reminders, with a $1,000+ private suit (FPF). (SB 942 / AB 853, live Aug 2, 2026, governs GenAI providers above 1M monthly users.)
3. CCPA ADMT & SB 951 Cal-WARN (Effective Jan 1, 2027): Significant-decision ADMT rules (Hunton) plus AI disclosures in covered mass-layoff notices.
4. SB 947 "No Robo Bosses" (Signed Sep 30, 2026; Effective Jul 1, 2027 per trackers): Bars sole ADS firing or discipline, requires human corroboration and notice, and sets a $500 fine per violation plus PAGA exposure (Reed Smith). See legal rules when a manager uses ChatGPT for layoffs or reviews.

• Colorado SB 26-189 (Signed May 14, 2026; Effective Jan 1, 2027): After xAI/DOJ suits and an April 27, 2026 stipulated pause (JD Supra), Colorado repealed its June 2026 law via SB 26-189 (Legal 500), requiring pre-decision notices, 30-day adverse reasons with human review, 3-year logs, and voiding vendor bias-indemnity shifts.
• NYC & Northeast: NYC Local Law 144 requires annual bias audits and 10-day notices (NYC DCWP), NY GBS Art. 47 governs companion bots, and secondary alerts note Connecticut SB 5 adds AI WARN disclosures Oct 1, 2026 and hiring notices Oct 1, 2027 (verify primary text with counsel).

Worked multistate math: A 45-person Boston firm hiring across MA, IL, TX, CA, and CO that runs an unvetted ATS scorer on 60 applicants (12 per state) without notices or logs violates Illinois HB 3773 on all 12 Illinois candidates regardless of intent, triggers California FEHA ADS agent liability and 4-year log rules, and faces Colorado SB 26-189 indemnity-voiding rules on Jan 1, 2027. One unified disclosure and a 4-year log spreadsheet take ~4 hours of HR time (~$240 at $60/hour) and $0 in software.

How does Massachusetts regulate workplace AI today without a dedicated AI Act?

Massachusetts regulates workplace, consumer, and contractor AI right now through general statutes carrying mandatory minimum and treble damages, even without a standalone "AI Act" as of October 29, 2026.

For silent meeting bots under s. 99, use our AI notetakers in meetings policy guide. Under s. 19B, Baker v. CVS Health (D. Mass., Feb 2024) let an AI video-interview claim survive dismissal before settlement, though later decisions narrowed s. 19B to deception tools (Chambers, May 21, 2026). Keep payroll PII out of consumer bots using what not to paste into an AI tool at work. At the Massachusetts State House, S.2619/H.5479 sit in conference and workplace monitoring bill S.35 remains in committee.

What changed in the EU AI Act after the July 2026 Digital Omnibus, and when does it reach a US firm?

For US firms serving Europe, Regulation (EU) 2026/1744 (published July 24, 2026; in force July 27, 2026: EUR-Lex; Garrigues) deferred Annex III high-risk HR and education obligations 16 months to December 2, 2027, softened Article 4 literacy wording, and kept Article 50 transparency rules live from August 2, 2026.

Top US search results mistakenly claim any US company with EU buyers or data is automatically covered. Under Article 2(1)(a)-(c) (Art. 2), a US-only firm enters scope only if it places an AI system on the EU market (2(1)(a)) OR "where the output produced by the AI system is used in the Union" (2(1)(c)). Internal AI use in Boston while billing a Munich client triggers GDPR, not the EU AI Act; running a support bot for French buyers or sending AI candidate scores to Dublin puts output in the Union.

Pair human editorial review under Article 50(4) with our source-checking and AI disclosure workflow, and satisfy Article 4 using our free 60-minute AI literacy session and scoreable AI literacy rubric.

Did your business insurance silently drop AI coverage at your 2026 renewal?

Commercial insurers spent 2026 stripping generative-AI losses from standard policies via three Verisk / ISO endorsements introduced in January 2026: CG 40 47, CG 40 48, and CG 35 08 (IIABA). State regulators approved over 80% of tracked carrier AI-exclusion filings (MoneyGeek).

• CG 40 47 & CG 40 48: CG 40 47 cuts Coverages A and B; CG 40 48 cuts Coverage B advertising/personal injury right as Anthropic's $1.5 billion Bartz settlement (482,460 books; 91.3% claim rate among eligible works) won July 2026 approval (Writer Beware).
• CG 35 08 & D&O/E&O Caps: CG 35 08 cuts products-completed operations; W. R. Berkley added absolute AI D&O/E&O exclusions (Jones Day), and $5M tech E&O policies may cap chatbot errors at $25,000 (Armilla). Under Moffatt v. Air Canada, firms stay liable for chatbot promises. Wire broker contacts into your first-hour AI incident response plan.

How should multistate employers plan for 2027? Three scenarios and the ed.cafe Scoreboard

Plan across three 14-month scenarios through December 2027: (1) State Patchwork Persistence (no federal statute; all state rules enforce), (2) Narrow First Amendment Injunctions (developer speech mandates trimmed; HR and wiretap rules untouched), or (3) Narrow Development Preemption (frontier training preempted; state civil-rights, fraud, privacy, and wiretap laws preserved). Under all three scenarios, employer workplace-AI duties survive.

Paste the ed.cafe Late-2026 State & EU AI Law Scoreboard & Employer Keep-Doing Checklist into your HR or operations channel. The ed.cafe Scoreboard puts every verified October 29, 2026 deadline and eight actions on one page.

THE ED.CAFE LATE-2026 STATE & EU AI LAW SCOREBOARD & EMPLOYER CHECKLIST
Verified: Oct 29, 2026 | Publisher: ed.cafe by No Worker Left Behind (EIN 88-3268312)
Operational status map for 10-to-500 person employers; not legal advice.

PART 1: FEDERAL PREEMPTION TRACKER
- EO 14365 (Dec 11, 2025): ACTIVE -> 0 state laws repealed.
- DOJ AI Task Force: LITIGATING (D. Colo.) -> 0 state HR/wiretap/fraud laws enjoined.
- FTC Draft Statement (Jul 1, 2026): PROPOSED -> 0 state HR or notice laws voided.
- Congress Moratorium: STALLED -> 99-1 Senate rejection (2025); Jun 2026 draft unpassed.

PART 2: JURISDICTION SCOREBOARD (OCTOBER 29, 2026)
1. ILLINOIS HB 3773 (IN FORCE Jan 1, 2026): Mandatory worker/applicant AI notice;
   no ZIP proxies; 4-yr records; draft IDHR Subpart J; IHRC fee-shifting.
2. TEXAS TRAIGA (IN FORCE Jan 1, 2026): Intent standard (disparate impact alone
   insufficient); AG-only, 60-day cure; $80K-$200K uncurable fines; NIST RMF defense.
3. CALIFORNIA STACK: FEHA ADS (IN FORCE Oct 1, 2025, 5+ staff, vendor=agent, 4-yr
   logs); SB 243 Chatbots (IN FORCE Jan 1, 2026, $1K+ private suit); CCPA ADMT +
   SB 951 Cal-WARN (EFF. Jan 1, 2027); SB 947 No Robo Bosses (SIGNED Sep 30, 2026,
   eff. Jul 1, 2027: human corroboration + notice, $500 fine + PAGA).
4. COLORADO SB 26-189 (SIGNED May 14, 2026; EFF. Jan 1, 2027): Replaced 2024 law.
   Pre-decision notice, 30-day adverse reasons + human review, 3-yr logs, voids
   vendor bias-indemnity shifts.
5. MASSACHUSETTS (IN FORCE VIA GENERAL STATUTES): Wiretap c. 272 s. 99 ($1K+ min);
   Lie Detector c. 149 s. 19B ($500+treble + app notice); 201 CMR 17.00 WISP (0
   floor); Ch. 93A/151B ($2.5M Earnest & $2.275M SafeRent); Policy AI.001 v1.1 for
   state vendors; S.2619 privacy bill in conference.
6. EU AI ACT + OMNIBUS REG. (EU) 2026/1744 (IN FORCE Jul 27, 2026): Covers US firms
   ONLY if placing AI in EU or output is used in EU (Art. 2). Art. 4 literacy LIVE;
   Art. 50 bot/deepfake rules LIVE Aug 2, 2026 (marking grace Dec 2, 2026); Annex III
   HR/education MOVED to Dec 2, 2027.
7. INSURANCE (SINCE Jan 2026): Spot ISO CGL exclusions CG 40 47, CG 40 48, CG 35 08,
   Absolute D&O AI exclusions, and $25K tech E&O chatbot sublimits.

PART 3: 8-ITEM EMPLOYER CHECKLIST
[ ] 1. Post one unified pre-decision AI notice on job ads & handbooks (HR).
[ ] 2. Require independent human corroboration + 4-yr selection logs (HR/Ops).
[ ] 3. Block unannounced external meeting bots & announce recordings (IT/Hosts).
[ ] 4. Verify MA s. 19B job-app disclaimer & 201 CMR 17.00 WISP rules (HR).
[ ] 5. Strip unverified "% accurate" or "replaces humans" ad copy (Marketing).
[ ] 6. Map EU-bound AI outputs & add Art. 50 bot labels before Dec 2, 2026 (Ops).
[ ] 7. Log a 60-minute role-based staff AI literacy session (Team Leads).
[ ] 8. Ask your broker in writing 45 days pre-renewal about CG 40 47, CG 40 48,
       CG 35 08, and E&O/D&O AI sublimits (Owner/CFO).

Through the Massachusetts AI Hub, the Google AI Professional Certificate is free for Massachusetts residents (Mass AI Hub; see is the Google AI Professional Certificate worth it). No Worker Left Behind (Boston 501(c)(3), EIN 88-3268312) is forming Circle #1 at ed.cafe/start/, a free 30-chair cohort closing December 1, 2026 (pre-launch as of October 2026; zero cohorts completed yet), alongside open guides in the Staff Room and Council page.

Where this stops being true

• Signed preemption or injunctions: Entries flip if Congress passes a preemption Act or a court enjoins a specific state law (watch D. Colo. after Jan 1, 2027 rulemaking).
• Draft rules and secondary alerts: Illinois Subpart J was reviewed in draft, Colorado AG rules are due Jan 1, 2027, and Connecticut SB 5 was checked via secondary legal alerts.
• Unions and regulated sectors: Unionized employers must bargain over AI under the NLRA (as in POLITICO's 2026 arbitration defeat: Allwork). Healthcare (case managers using AI for case notes) and finance carry stricter rules.
• Not legal or brokerage advice: Retain licensed counsel or a broker upon any agency inquiry, worker claim, or CG 40 47 renewal notice.

Questions people ask

Key terms

• Automated-Decision System (ADS / ADMT): Software or AI that scores, ranks, filters, or materially influences hiring, promotion, pay, discipline, or termination decisions.
• Disparate impact vs intent standard: Split where Illinois and California penalize neutral AI tools that disproportionately screen out protected groups, whereas Texas TRAIGA requires intentional discrimination.
• Digital Omnibus on AI (Reg. (EU) 2026/1744): July 2026 EU regulation softening Article 4 literacy language and deferring Annex III high-risk AI rules to December 2, 2027.
• Extraterritorial output test (Art. 2(1)(c)): EU AI Act rule covering non-EU firms when output produced by their AI system is used inside the European Union.
• Federal preemption: Constitutional displacement of conflicting state law by an Act of Congress or statutorily authorized federal regulation.
• ISO CG 40 47 / CG 40 48 / CG 35 08: Verisk commercial general liability endorsements available since January 2026 that exclude generative-AI claims.

Sources

1. Sidley Austin, "Dec 11, 2025 Executive Order on AI" (Dec 23, 2025): Sidley
2. Proskauer, "DOJ Intervenes in Colorado AI Case" (Apr 28, 2026): Proskauer
3. FTC, "Policy Statement on AI Accuracy" (Jul 1, 2026): FTC
4. CASRAI, "Federal AI Moratorium Fight" (Sep 20, 2026): CASRAI
5. Hinshaw, "Illinois HB 3773 Employment AI Rules" (Feb 26, 2026): Hinshaw
6. Latham & Watkins, "Texas TRAIGA" (Jun 2025): Latham
7. Mayer Brown & Reed Smith, "CA FEHA ADS & SB 947": Mayer Brown and Reed Smith
8. Legal 500, "Colorado SB 26-189 Overhaul" (Jun 18, 2026): Legal 500
9. Mass.gov & MA Legislature, "s. 99, Earnest $2.5M Settlement & S.2619": s. 99, Earnest, and S.2619
10. Official Journal of the EU, "Regulation (EU) 2026/1744" (Jul 24, 2026): EUR-Lex
11. IIABA & MoneyGeek, "ISO CG 40 47/40 48/35 08 Exclusions": IIABA and MoneyGeek
12. IBM Newsroom, "2026 Cost of a Data Breach (n = 602)" (Jul 29, 2026): IBM