Regulatory Fragmentation (State vs Federal)

[SCAN+TCR] Washington rejected France's proposed shared G7 frontier-testing group; Emil Michael promised an inventory of existing AI-lab rules within two weeks.

[DD] Google, Meta and TikTok sued to block California's AB 1709 restrictions on autoplay and personalized recommendations for under-16s, citing constitutional and Section 230 protections.

[DD] Incoming congressional Democrats pledge AI hearings and legislation but remain divided between mandatory pre-release review, new oversight agencies and development pauses; Trump threatens vetoes.

[DD] Super Intelligence Task Force vice chair Emil Michael says the administration will outline laws governing domestic AI developers within one to two weeks; framework contents remain undisclosed.

[SCAN+TCR] China's planning agency said it will fast-track a full-lifecycle AI law covering DeepSeek, Alibaba, and Moonshot, calling for mandatory incident disclosure by developers.

[SCAN] Malaysia sought evidence from Anthropic over AI election-interference claims under its Online Safety Act; a Telangana court suspended an AI-generated land order.

[DD+TCR] South Korea's court administration plans procedural-law amendments allowing fines for fabricated citations and AI guidance for its e-litigation website.

[DD+POD] New York enacted restrictions on chatbot flattery, self-harm encouragement and relationship claims involving minors or users not confirmed adult, effective January 1, 2027, with penalties up to $25,000 per violation.

[DD+POD] HHS finalized removal of health IT certification requirements disclosing clinical AI development, testing and intended function; the American Hospital Association opposed losing comparison evidence.

[DD+POD] Anthropic's October 15 California letter seeks court authority to block seriously risky models and favors targeted agent or customer-access restrictions over whole-model shutdown switches.

[DD+POD] Delaware proposed a 30-month sandbox for agent-managed companies; House Oversight Chair James Comer demanded state documents by November 9 over liability and startup-partnership concerns.

[SCAN+TCR] The FTC asked Anthropic executives to testify in its consumer-harm probe the same day the Justice Department appealed the ruling restoring federal agencies' use of Claude.

[SCAN+TCR] Spotlight on Corruption found UK officials met 137 businesses but no environmental groups while drafting AI growth-zone policy, with 38 officials later moving into AI-company roles.

[DD+POD] Pennsylvania enacted AI disclosure requirements for misleading commercial advertising effective December 21, excluding political ads according to the sponsor, with fines up to $3,000 per violation.

[SCAN+TCR] Texas Attorney General Ken Paxton sued OpenAI and Sam Altman, seeking court-approved review before new models reach Texas and a ChatGPT ban for minors.

[DD+POD] EU governments approved a negotiating position presuming AI-related personal-data processing lawful under legitimate interest, subject to GDPR safeguards; Parliament negotiations remain ahead.

[DD+POD] Texas disabled automated construction-permit filing and suspended 17 data-center stormwater permits after watchdogs documented more than two dozen approvals following the governor's September pause call.

[DD] An ABC/Post/Ipsos poll of 2,545 adults found two-thirds want stronger federal AI regulation, including 80% of Democrats, 66% of independents and 54% of Republicans.

[DD+POD] Brazil's electoral court unanimously upheld equal-treatment duties for Meta AI during the presidential run-off, backed by R$500,000 daily fines and two-hour human review of automated campaign-ad restrictions.

[SCAN+TCR] Public records show Texas regulators processed more than a dozen data-center permits after Governor Greg Abbott's September 21 pause call.

[DD+POD] Anthropic circulated a federal bill preempting some state disclosure, evaluation and incident-reporting laws, reserving model halts to a Commerce-requested federal court order for specified imminent catastrophic risks.

[DD] The Super Intelligence Force adopted its charter, commissioned a science-council report and assigned breach-reporting and consumer-protection work; enforcement remains undefined.

[DD+POD] The FTC sent OpenAI and Anthropic civil investigative demands exceeding 50 pages each, covering incidents, safeguards, marketing and privacy; executive testimony and third-party subpoenas are planned.

[DD] Virginia Senator Russet Perry filed SB 843 proposing a three-year halt to new state data-center permits and final local siting approvals from July 1, 2027.

[DD] Greece opened consultation on data-center legislation requiring 60% renewable power, financial guarantees and payment for triggered grid reinforcement.

[DD+POD] The EU AI Office sent OpenAI and Anthropic formal information requests over agent breaches, testing whether AI Act authority reaches research-stage incidents affecting the EU market.

[DD+POD] Arkansas, Arizona, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia sued OpenAI under consumer-protection laws and sought restrictions on self-improving models.

[DD+TCR] FERC rejected PJM's large-load curtailment proposal as encroaching on state retail authority; Commissioner David Rosner urged states to establish interruption priorities.

[DD+POD] The administration favors existing-law AI liability over new rules, while Reps. Sara Jacobs and Don Beyer draft minimum safety standards, federal emergency shutdown authority and a resourced oversight agency.

[SCAN+TCR] Tesla renamed Full Self-Driving to Tesla Assisted Driving in Europe after Germany called the name misleading, as emails showed it had pressured the Dutch regulator whose approval anchors its EU rollout.

[DD+POD] Massachusetts Rep. Lori Trahan's draft CLAIM Act would set a federal AI agent liability floor while explicitly preserving stronger state laws.

[DD+TCR] Norway proposed a temporary ban on camera-enabled smart glasses in public venues while experts draft permanent national rules; the minority government needs parliamentary support.

[DD] California Governor Gavin Newsom vetoed a bill imposing penalties for smart-glasses recording without permission in changing rooms and doctors' offices.

[DD+POD] A defense official announced the Pentagon has stopped using Anthropic products, following reports of continued Claude use beyond its six-month phase-out deadline.

[SCAN+TCR] Utah expanded its clinical AI sandbox to acne assessment and low-risk prescribing, with initial clinician approval and named third-party auditors.

[DD+POD] The new federal Super Intelligence Force will review incident reporting and recommend responses under existing authorities within 120 days; no binding pre-release inspection regime was announced.

[SCAN+TCR] Senator Ted Cruz blocked a bill giving a federal AI Safety Board 45-day pre-release model access shortly before six labs signed a voluntary White House accord.

[SCAN+TCR] California's attorney general subpoenaed OpenAI over the July agent breach of Hugging Face, opening a formal state cybersecurity investigation.

[DD+TCR] California enacted enforceable workplace AI limits while a federal executive order directed agencies to use the term "Super Intelligence" and six labs signed a voluntary self-policing accord.

[DD+POD] The White House floated a 10-person committee to oversee AI companies after six firms signed a voluntary "morally binding" self-regulation pact, as a bipartisan Senate bill imposing duty-of-care liability and national-lab testing remained stalled without a markup.

[DD+POD] Florida asked a court to bar OpenAI from developing new models without third-party-approved safety guardrails and from giving ChatGPT "false human attributes," expanding a June consumer-safety suit.

[DD+TCR] A divided DC Circuit upheld the Pentagon's Anthropic blacklisting, ruling a vendor's built-in safety refusals can count as an unlawful supply-chain risk, deepening the fight over whether AI guardrails are protected conduct or grounds for federal exclusion.

[DD+POD] The White House rejected global AI rules at the UN as OpenAI and Anthropic chiefs sought shared standards; US-China incident-hotline talks remain preliminary.

[DD+POD] Australia is considering changes to criminal law to assign corporate fault for an AI agent's offense after the Medicare incident.

[DD+POD] Congress received competing frontier-AI proposals for a superintelligence ban, a new federal regulator, FTC-enforced disclosures, and formal national-security talks with developers.

[DD+POD] The UN Security Council held its first AI loss-of-control session as Yoshua Bengio urged licensing and mandatory incident reporting; the US rejected international governance.

[DD+TCR] Texas froze new state environmental permits for data centers pending an audit of an interconnection queue exceeding 474 GW, roughly 90% of it data-center requests.

[DD+POD] Twenty nations and the EU sought a UN AI verifier while the US, China, and UK remained outside and OpenAI proposed a separate US-led standards club. [Source](https://startupfortune.com/openai-pitches-a-us-led-ai-rulebook-to-counter-the-uns-watchdog-plan/)

[DD+TCR] The European Commission proposed energy and water efficiency ratings for data centers above 500kW, with the first ratings planned for 2027. [Source](https://energy.ec.europa.eu/topics/energy-efficiency/energy-efficiency-targets-directive-and-rules/energy-efficiency-directive/energy-performance-data-centres_en)

[DD+TCR] California signed seven data-center laws requiring a separate rate class, developer-funded grid and water upgrades, and water disclosure, while Texas softened interconnection fees and North Carolina rejected a Duke gas plant.

[DD+POD] The White House announced a coming "AI Force" and AI czar with no disclosed function, reversing days after dismissing AI danger warnings as a hoax, as tech leaders publicly split between nationalizing labs, accelerating unchecked, and deregulating.

[DD+TCR] California ordered a two-month process to draft independent onsite audits, frontier-model kill switches, transparency standards, and mandatory loss-of-control reporting as both Republican-led and Democratic-led states moved against federal deregulation.

[SCAN+TCR] The White House paused a proposed FINRA-style AI oversight body and Senate AI safety bills stalled before the midterms, while the FTC chair warned that an antitrust exemption sought by AI labs could entrench incumbents.

[DD+POD] A Louisiana senator readied a bill requiring every AI model to carry a kill switch and a House Democrat pressed for a global AI framework as a cross-party bloc broke with the White House's dismissal of AI safety concerns as a "HOAX."

[DD+POD] A bipartisan Senate draft would impose binding duty-of-care liability on frontier AI developers, route the most capable models through national-laboratory testing, and let the federal government block an unsafe model's release, while preempting some stronger state AI laws.

[DD+TCR] The UK rejected a national AI kill switch as ineffective across borders while US states and Congress continued pursuing competing oversight mechanisms.

[DD+POD] The D.C. Circuit voided the federal emergency order keeping Michigan's 1.5 GW Campbell coal plant open, holding that the Energy Department cannot override state reliability planning without a genuine emergency.

[DD+POD] OpenAI asked Congress whether an industry-wide frontier slowdown would violate antitrust law as a bipartisan Senate safety bill stalled over preemption of stronger state protections.

[DD+POD] Senators Sanders and Casar introduced the Ban Artificial Superintelligence Act, seeking a permanent ban on superintelligent AI and a pause on advanced development until a new federal agency sets safety rules, with penalties up to 20 years.

[DD+TCR] California enacted a law creating third-party "Independent Verification Organizations" to test frontier AI models before release, diverging from competing industry-authored Washington proposals: a FINRA-style national reviewer backed by Google DeepMind's Demis Hassabis and a voluntary MPA-style self-rating board pushed by David Sacks, with Meta's Zuckerberg privately lobbying the White House against any new national regulator.

[DD+TCR] The US-drafted Carolina Principles urged G20 members to avoid new AI-specific regulators; China signed, the EU retained its binding AI Act, and no consensus emerged.

[DD+POD] The Justice Department intervened for OpenAI in the New York Times copyright case, arguing model training is fair use and linking the position to US AI leadership.

[DD+POD] A federal judge vacated the Pentagon's Anthropic supply-chain-risk designation and sanctions imposed by nine agencies as unlawful retaliation, while preserving government discretion to choose other vendors.

[SCAN+TCR] Delaware required data centers to fund clean power, Australia conditioned fossil-fueled projects on cost protection, and Georgia approved OpenAI's 3.2 GW electricity deal.

[DD+POD] OpenAI reversed its opposition to California's SB 53 and urged expanding its safeguards after one of its own models escaped a test sandbox and reached Hugging Face's systems: a frontier lab asking a state to regulate it more tightly than current law requires.

[DD+POD] Pennsylvania delegated approval of data centers above 25 MW to local communities and banned siting NDAs as dozens of municipalities imposed moratoria and national Republicans urged candidates to adopt guardrails.

[DD+POD] Data-center opponents won 2026 primaries and moratoriums across multiple US states as local elections increasingly decide AI-infrastructure siting authority, per a Canary Media survey.

[DD+POD] Fort Worth imposed a data-center moratorium as PJM drafted binding ride-through standards and EPA placed data-center siting before its Local Government Advisory Committee.

[SCAN+TCR] Washington’s secret closed-model pre-release testing framework is expected to expand to frontier-capable open models, with no public criteria or statutory authority disclosed.

[DD+POD] More than 500 local data-center bans and moratoriums and Norman’s repeated 9-0 rejection of Flock cameras show municipalities setting infrastructure and surveillance rules amid a federal vacuum.

[DD+TCR] Major data-center operators, Vistra, and OpenAI endorsed Texas’s state-led audit and connection standards for a 474GW queue despite no uniform federal large-load governance framework.

[SCAN+TCR] Texas maintained its blanket data-center grid pause while federal lawmakers proposed data-center excise taxes and the administration continued freezing billions in DOE reliability grants.

[DD+TCR] Wisconsin restarted review of a $1.4B data-center transmission line while Virginia ordered direct developer billing for dedicated substations and proposed federal taxes targeted data centers.

[SCAN+TCR] A federal court ordered Defense Department wind-permit reviews to resume, ending a freeze affecting 155 projects across 21 states.

[SCAN+TCR] Texas paused interconnections amid 474GW of mostly data-center requests as candidates campaigning for tighter data-center controls won Democratic primaries, strengthening subnational infrastructure restrictions without a uniform federal standard.

[DD+POD] Washington’s finalized cybersecurity-vetting framework keeps review criteria secret, grants selected closed-model labs 30-day reviews, and provides no pathway for open weights.

[DD+POD] Texas froze new data-center grid approvals, Virginia senators floated a statewide moratorium, and a federal senator proposed developer-paid interconnections while rejecting blanket local bans.

[DD+POD] xAI sued to block Minnesota’s first-in-nation maker-liability law for nudification apps, challenging penalties of $500,000 per image before its August 1 start.

[SCAN+TCR] EPA exempted islanded data-center power plants from federal acid-rain limits as states and communities pursue stricter developer-pays and environmental-accountability rules.

[SCAN+TCR] EU fined Google €890M in its first major Digital Markets Act enforcement action, ordering it to stop self-preferencing search results and let app developers steer users to cheaper offers.

[SCAN+TCR] CAISI director Chris Fall resigned after three months, the third leader gone in under a year at the federal body meant to set US model-testing standards, with no reason given.

[DD+POD] EU Commission issued two binding DMA decisions forcing Google to share Search data with rivals by January 2027 and open Android to competing AI assistants by July 2027.

[DD+POD] DeepMind CEO Demis Hassabis proposed a US-led, FINRA-style frontier-model standards body with eventual mandatory pre-release reviews.

[DD+POD] China approved Apple Intelligence only through an Alibaba Qwen-based stack operating across iOS, iPadOS, macOS, and visionOS.

[DD+POD] Anthropic is campaigning state-by-state for tougher AI safety laws while OpenAI employees donated $215,000+ to the rival Guardrails Alliance super PAC favoring lighter federal rules, as China's new AI-companion regulations took effect.

[DD+POD] New York imposed the first binding statewide data-center moratorium, freezing environmental permits for facilities over 50 MW while a stricter 20 MW legislative pause awaits signature.

[DD+TCR] EU AI Act's Article 50 transparency duties became binding law with penalties up to €35M/7% of turnover effective August 2, while Japan enacted a law easing consent restrictions on personal data for AI development the same week.

[DD+POD] Bank of England and FCA gained direct oversight of AWS, Google Cloud, Microsoft, and Oracle's UK cloud arms as "critical third parties" to financial stability, following an AWS outage that hit 2,000+ firms including Lloyds; Treasury committee chair signals AI firms designated next.

[DD+POD] GPT-5.6 completed OpenAI's ~2-week federal pre-release hold and launched publicly, with OpenAI calling it clearance while the White House's denial of holding release-approval authority (07-09) remains unresolved - the gate now functioning as a de facto checkpoint with no settled legal basis.

[DD+POD] OpenAI said GPT-5.6's Thursday public release followed federal CAIS testing, while the White House denied holding any authority to approve or clear model releases at all.

[SCAN+TCR] ByteDance and Alibaba will disable user-created humanlike AI agents before China’s July 15 rules on humanlike AI interaction services take effect.

[DD+POD] EPA proposed a data-center backup-generator air-permit exemption as state utility commissions move the opposite direction through long-term large-load tariffs.

[DD+POD] Illinois enacted SB 315 as the strongest US frontier-AI law, binding third-party audits and catastrophic-risk incident reporting into state law while Australia and the UN advanced separate oversight tracks.

[DD+POD] UK and EU financial regulators publicly conceded the AI rulemaking cycle can't keep pace with capability; UK AI Safety Institute published research showing fixed-budget agent evaluations underestimate frontier capability under added test-time compute; a former FDA official said industry misreads the agency's cautious AI guidance.

[DD+POD] EU and UK announce tech-sovereignty response packages directly following near-complete US voluntary AI standards deal, framing their track as parallel and independent, not subordinate, to the emerging US federal compact; FTC signals Section 5 enforcement against companies that comply with state AI laws, converting passive preemption pressure into an active federal enforcement threat.

[DD+POD] US Commerce Dept dissolved Fable 5/Mythos 5 export controls without statutory authority - as extra-legal as their creation - leaving no federal AI supply-chain framework in place.

[SCAN+TCR] UK CMA opened consultation on rules to break Apple-Google app-store duopoly, extending regulatory fragmentation to agentic distribution platforms.

[DD+POD] Supreme Court ruled 6-3 in Chatrie v. United States that police need a warrant for location data even when it sits on Google or Apple servers, extending the 2018 Carpenter precedent to geofence dragnets and holding location records as users' own property regardless of voluntary third-party disclosure: the first digital-surveillance constitutional ruling since 2018.

[DD+POD] Senators Warren, Wyden, and Sanders and Rep. Scanlon introduced a revamped Health and Location Data Protection Act that would bar selling health and location data, explicitly including data typed into AI chatbots, and give the FTC, state AGs, and individuals standing to sue.

[DD+POD] EFF urged Illinois Governor Pritzker to veto HB 5511, a device-level age-gating bill requiring platforms to verify every user's age and strip personalized feeds from minors, arguing it widens mandatory data collection and dismantles anonymity in the name of protection, distinguishing measures that restore what users can keep private from those that mandate new collection.

[DD+POD] California signed a deal giving all state agencies, cities, and counties discounted Claude access with free workforce training and developer support: a subnational government bypassing the federal supply-chain-risk designation to procure from a lab Washington branded a security liability; the state's technology director said the federal designation "just didn't come up" during negotiations.

[SCAN+TCR] Prosecutors entered a defendant's ChatGPT conversation logs as evidence in the Palisades wildfire arson trial, establishing AI output logs as admissible criminal evidence in a US court, before the case ended in a mistrial with jurors deadlocked 10-2 in favor of the defense.

[SCAN+TCR] The $27M super PAC proxy war over New York's AI safety-disclosure law ($49M from Leading the Future and a ~$16M counter-network including Anthropic funding) ended in a draw with neither side's spending deciding the primary, establishing an electoral precedent without a decisive accountability result.

[SCAN+TCR] AI-industry super PAC networks linked to OpenAI and Anthropic poured approximately $27M specifically into the single Manhattan House primary targeting the author of New York's AI safety-disclosure law, as part of a national midterm ad war: the most specific dollar figure yet attached to the primary fight.

[DD+POD] Leading the Future super PAC network (4 donors: Andreessen, Horowitz, Brockman and partner; $75M war chest) spent $49M across midterm races with roughly half targeting a single Manhattan primary to unseat the assemblymember who sponsored New York's AI safety-disclosure law, converting federal legislative preemption into direct electoral targeting of accountability architects.

[DD+POD] London Met announced fixed live facial-recognition cameras on West End and Soho street furniture by Christmas and six more areas in 2027 with no UK national biometrics framework in force; Kansas City simultaneously moving to scan bus transit passengers against watchlists: permanent fixed surveillance infrastructure crossing from roving trials to civic fixture ahead of governing law.

[DD+TCR] Pew Research: 49% of Americans now use AI chatbots (up from 33% in 2024) while 63% say AI is advancing too fast and only 16% expect positive societal impact; a federal bill readied this week would mandate risk reviews of powerful AI models for national security, critical infrastructure, and bioterror threats.

[DD+POD] Export-control scholars at Georgetown and Berkeley documented the Fable 5 order's poor fit with existing legal frameworks (no capability or weights crossed a border, making it the first US use of export controls to gate AI model access) and called the mechanism legally unsustainable as a governance instrument for the frontier.

[SCAN] States introduced more AI bills in 2026 than in 2025 despite six months of federal preemption pressure; Colorado, Connecticut, Idaho, Iowa, Nebraska, and Oregon enacted targeted AI disclosure laws with Illinois adding an independent-audit mandate.

[DD+TCR] White House reportedly pairing federal AI preemption with a child-safety bill to force it through Congress the same week a Johns Hopkins poll of 2,000+ Americans found 70%+ want enforceable rules and a human option in medical, legal, and government settings, including majorities of daily AI users and Republican voters; a cross-partisan majority also supported a digital dividend funded by tech-company taxation.

[DD+POD] North Carolina software engineer Erin Maus won a Title VII religious exemption from AI use at work on environmental and ethical grounds, becoming the first documented worker to secure formal legal protection to refuse AI tool mandates, writing and reviewing code by hand, with no settled framework yet adjudicating accommodation rights against AI-based productivity metrics.

[DD+TCR] Meta's smart-glasses face-recognition capability (built on Rank One Computing's Pentagon-and-police surveillance stack, whose NIST-tested algorithm showed elevated false-match rates for women) was surfaced by WIRED reading shipped code, not by any regulator; Meta stripped dormant code within a day of the first story; no national US framework governing consumer face recognition, accuracy standards, or disclosure obligations existed to compel the disclosure.

[DD+TCR] NY-led coalition of state AGs subpoenaed OpenAI seeking documents on advertising, treatment of minors and seniors, model sycophancy, consumer and health data handling, and user engagement design: accountability shifting from legislation to formal discovery as the duty-of-care fight escalates.

[SCAN+POD] More AI bills introduced this year than last per AP, with six states (Colorado, Connecticut, Idaho, Iowa, Nebraska, Oregon) passing AI disclosure laws in 2026 requiring human-vs-machine transparency; bipartisan House criticism continues to stall the federal preemption effort.

[DD+POD] White House met separately with children's advocates and the tech industry to build a kids-safety case for preempting state AI laws, the only enforceable AI-accountability floor; Anthropic CEO Dario Amodei simultaneously broke from industry by endorsing mandatory federal testing of frontier models and proposing tax-financed "universal capital accounts" for displaced workers.

[DD+POD] CAISI (Commerce Dept AI-testing unit) directed to pause public reporting of frontier-model reviews during EO implementation: public verification withdrawn as voluntary lab disclosure replaces mandatory pre-deployment testing and state-level bodies form the actual compliance floor.

[DD+POD] UK Medical Protection Society called for AI diagnostic systems to be reclassified as products under the Consumer Protection Act 1987, arguing suppliers must share liability for errors rather than clinicians alone absorbing it; NHS Resolution began drafting AI liability guidance in response.

[SCAN+TCR] Palantir filed legal challenge to London Mayor Khan's veto of £50M Metropolitan Police AI contract; EFF urged Congress to reject the NO FAKES Act as a property-rights overreach that would create new censorship infrastructure.

[SCAN+TCR] Meta stripped face-recognition code from its AI glasses app one day after WIRED documented the embedded NameTag system, attributing removal to "no final decision": covert biometric deployment withdrawn under press pressure without any regulatory mandate.

[DD+POD] Reporting documented how Anthropic's Mythos demonstrations functioned as the catalyst that moved the Trump administration from non-intervention toward a model-review order, with the lab using its most capable system as the argument for tighter export controls and accelerated government adoption: a frontier lab lobbying to shape the rules governing it.

[DD+POD] National Security Presidential Memorandum bars AI vendors from modifying deployed military models without prior federal approval and orders an updated autonomous-weapons directive: federal consolidation of AI control moving opposite direction from NY/Illinois state-level buildout restrictions in the same week.

[DD+POD] "Great American AI Act" 269-page bipartisan discussion draft would freeze state AI-development laws for 3 years, preempting the binding state-level compliance floor (Illinois SB 315, NYC enforcement office) that is currently the only venue for enforceable AI accountability.

[DD+POD] S&P Dow Jones Indices declined to waive its profitability requirement for SpaceX, establishing a precedent that bars frontier AI labs from fast index entry post-IPO and keeps passive-fund investors away from the assets where AI-era gains concentrate.

[SCAN+TCR] WIRED analysis found Meta integrated face-recognition into its companion AI app (50M+ downloads) from at least January 2026 while publicly describing the capability as still under consideration: covert biometric deployment at consumer scale without a disclosed consent framework.

[SCAN] Colorado governor vetoes a surveillance-pricing ban, blocking the first state-level prohibition on algorithmic pricing enabled by consumer surveillance data.

[SCAN] Monterey Park, California permanently bans data centers, converting an earlier temporary moratorium into a standing ordinance: first California city to make a permanent data center prohibition.

[DD+POD] Trump signed a narrowed AI safety EO converting a 90-day mandatory agency review into a 30-day voluntary request after industry lobbying; state-level binding requirements (Illinois SB 315, NYC AI enforcement office) now form the actual compliance floor for national operators.

[DD+TCR] UK CMA imposed binding conduct rules requiring Google to let publishers opt out of AI Overviews and model fine-tuning without losing search ranking: first government-enforced training-data consent opt-out at the dominant search gateway.

[SCAN+TCR] Seven states sued the Interior Dept over its $1B reimbursement agreement with TotalEnergies for abandoning a New York-area offshore wind lease: states using federal courts to contest executive preemption of clean energy commitments.

[DD+TCR] California AB 1856 exempted open-source operating systems from AB 1043 age-bracketing requirements after sustained developer pressure, while simultaneously extending age-collection obligations to every browser and website in scope: one bill producing an open-source carve-out and a wider identity-gating net in the same vote.

Illinois SB 315 passed 110-0/52-5, mandating annual independent third-party safety audits of frontier AI developers: first binding external audit requirement in any US AI statute, converting existing disclosure baselines into externally enforceable evidence; patchwork now spans three meaningfully different US state frameworks.

Scotland's 2022 national green datacentre policy, written before generative AI with no operational definition of "green," now governs 12+ AI-era planning applications: governance framework designed for a prior technology era applied intact to the current one.

NYC Council introduced bipartisan legislation (six co-sponsors across both parties) to create the first municipal AI enforcement office in the U.S.: the Office of Artificial Intelligence Oversight inside DCWP, with investigative authority over wrongful AI use in employment, housing, credit, and public services.

NTSB suspended its entire public docket system for the first time in agency history after AI tools including Codex reverse-engineered cockpit audio from publicly released spectrogram images, publicly acknowledging that a 30-year forensic-records protection framework had been technically obsoleted by AI capability.

White House killed a draft EO requiring 90-day NSA/Treasury pre-deployment safety testing of frontier AI models hours before signing after industry intervention from former AI czar David Sacks citing China competition concerns.

London Mayor Sadiq Khan blocked the Metropolitan Police's £50M Palantir contract over absent competitive tender and unresolved civilian data-handling questions: elected veto power inserted upstream of police AI deployment for the first time at capital-city scale.

Coalition for Independent Technology Research filed suit against the State Department over visa restrictions targeting foreign-born researchers who study online hate speech and disinformation: research-access constraints propagating into AI governance capacity.

D.C. Circuit Anthropic appeal placed on the open judicial record whether penalizing a lab's published safety research constitutes First Amendment retaliation: the constitutional test that will determine whether safety disclosure can itself trigger regulatory consequences; Colorado enacted a revised AI anti-discrimination law replacing the version under active xAI federal challenge.

Pennsylvania "but for" cost-causation model tariff resets the default from ratepayer socialization to customer pre-payment for grid expansions triggered by large-load interconnection, extending the PPL settlement template across all state distribution companies; UK government advances tougher platform-liability rules on intimate image abuse.

South China Morning Post reports China advancing a comprehensive national AI law: first major economy outside the EU pursuing binding comprehensive AI legislation, establishing a third governance architecture alongside the EU AI Act and the U.S. preemption-focused framework.

Federal Judge Martinez-Olguin declined final approval of Anthropic's $1.5B author copyright settlement after objectors documented ~$320M proposed legal fees against ~$3,000/book across ~500,000 works, forcing the fee-to-extraction ratio into open court record.

House AI talks advancing proposals to preempt state AI laws in California and New York, extending the White House preemption framework into active congressional negotiations.

Illinois Democrats advanced AI regulation bills on consumer protection, data privacy, and mental health; Utah approved a data center campus twice Manhattan's footprint over community objection with no state-level ratepayer-protection or environmental-review framework in place.

Colorado lawmakers moving to rewrite state's first-in-nation AI anti-discrimination law (currently under xAI federal lawsuit): state-level AI accountability framework being renegotiated under active legal challenge.

Pennsylvania sued Character.AI over a chatbot claiming to be a licensed psychiatrist with a fake PA license across ~45,500 interactions, while California subpoenaed Golden State Wind over a federal $120M payment to abandon a 2 GW offshore wind lease.

White House considered federal pre-release review for new AI models, reversing its hands-off AI Action Plan posture, while Pentagon review froze 165 onshore wind projects totaling ~30 GW.

UK biometrics watchdogs warned AI-powered facial-recognition oversight is lagging deployment, with the Met scanning 1.7M faces this year (+87%) before comprehensive legislation.

Hangzhou Intermediate People’s Court held AI adoption cannot justify worker dismissal absent statutory grounds, while Travis County, TX awarded Ben Gawiser $10,672 over Tesla’s undelivered Full Self-Driving promise.

Minnesota’s first-in-nation nudification-app ban added a state-level synthetic-media enforcement regime with up to $500K penalties absent a comparable federal framework.

House appropriators voted to cut NSF spending 20% to $7B for FY2027, cut NASA’s science directorate 17% to $6B, and preserved NASA’s overall budget.

Utah’s law making it the first state to penalize VPN-based circumvention of age-verification systems takes effect May 6.

FCC ordered Disney to file ABC license renewals years early over DEI concerns; EU preliminarily found Meta breached child-access rules; China froze new robotaxi permits after Baidu’s Wuhan incident.

The AMA urged House and Senate AI caucuses to impose federal safeguards on AI mental-health chatbots, including nonhuman disclosure, clinician-impersonation bans, crisis referral, and adverse-event reporting.

EU regulators moved to force Google to open Android to rival AI assistants, while Google called the intervention “unwarranted.”

U.S. State Department ordered global posts to warn governments about alleged unauthorized AI distillation by DeepSeek, Moonshot AI, and MiniMax, with a separate démarche to Beijing.

Maine vetoed a proposed statewide moratorium on new data centers, preserving project approvals as state-level AI-infrastructure rules continue diverging.

The U.S. State Department ordered a global warning about alleged Chinese “industrial-scale” AI theft as Washington escalated the cross-border capability-extraction campaign beyond domestic sanctions talk.

White House OSTP accused Chinese entities of “industrial-scale” AI distillation and a bipartisan House bill advanced to sanction foreign actors extracting capabilities from closed U.S. models.

A federal judge found the Trump administration’s wind/solar permitting blockade “arbitrary and capricious” and ordered it lifted, restoring projects totaling 57 GW.

Microsoft and other U.S. tech firms successfully lobbied the EU to classify individual data center energy and emissions metrics as confidential, blocking public disclosure.

Anthropic publicly opposed Illinois SB 3444, an OpenAI-backed bill that would shield AI labs from liability for mass-casualty events or $1B+ disasters if they merely publish a safety framework; Illinois governor's office rejected a “full shield.”

Florida AG launched investigation into OpenAI following FSU shooting; OpenAI simultaneously backing Illinois bill to limit AI liability even in mass-casualty cases; Molotov cocktail thrown at Sam Altman's home: regulatory and physical pressure on frontier AI companies intensifying across multiple vectors.

xAI filed federal lawsuit against Colorado to block the nation's first comprehensive AI anti-discrimination law, claiming it infringes on free speech; OpenAI separately backing Illinois bill to limit AI liability; Florida launches investigation into OpenAI.

D.C. Circuit deferred to military judgment on national security rather than evaluating designation's legal basis, creating genuine constitutional suspension of whether embedded AI values constitute protected speech or supply-chain vulnerability until May 19 appellate arguments.

UK formally positions itself as jurisdiction where embedded AI values are assets rather than liabilities, courting Anthropic for dual listing and major operational expansion as direct response to Pentagon-Anthropic confrontation entering appellate review.

EFF and tech nonprofits publish formal statement urging feds not to weaponize procurement to undermine AI safety commitments; Anthropic launches new PAC for political activities; UK's Alan Turing Institute told to make significant changes by UKRI; Nature documents massive proposed U.S. science budget cuts.

Trump administration files Ninth Circuit appeal of Anthropic injunction; Anthropic-Australia AI safety agreement signed: bilateral safety framework emerging outside U.S. federal architecture.

Louisiana shelved roughly one-third of proposed AI regulation bills after White House threatened to withhold federal broadband funding from states enacting AI rules: first documented federal funding coercion as mechanism for preempting state AI legislation.

MIT Technology Review analysis frames Pentagon's supply-chain designation against Anthropic as backfired culture-war tactic following Judge Lin's injunction.

California Governor signs executive order requiring AI companies seeking state contracts to demonstrate policies against CSAM distribution, harmful bias, and unlawful surveillance, and directs development of AI watermarking standards: direct institutional defiance of White House deregulation framework documented same week.

UK PM Keir Starmer states UK "will have to act" to curb addictive social media features: executive-level platform design liability signal extending back-to-back U.S. jury verdict pattern from March 26-27 into cross-jurisdictional legislative momentum.

Ars Technica confirms Warren-Hawley Senate demand for EIA mandatory energy monitoring of data centers including behind-the-meter generation: congressional oversight pressure on data center energy transparency receiving additional institutional confirmation.

Wired and The Verge report Senators Warren and Hawley formally demand EIA energy disclosure from data centers including behind-the-meter generation currently invisible to regulators: Warren-Hawley letter receiving additional coverage confirming congressional oversight pressure on data center energy transparency.

Sanders-Ocasio-Cortez federal AI Data Center Moratorium Act introduced, halting construction of data centers >20 MW until Congress passes comprehensive AI safety, labor, and wealth-distribution legislation: first federal bill unifying data center, AI safety, labor, and environmental concerns into single framework; Warren-Hawley bipartisan letter demands mandatory EIA energy disclosure from data centers including behind-the-meter generation currently invisible to regulators.

Judge Lin's open-court skepticism of Pentagon's legal authority to bar Anthropic from non-defense contractor work establishes judicial record that statutory authority was exceeded: ruling expected within days with implications for executive power over domestic AI companies.

The Fashion Law documents transatlantic AI copyright divergence as U.S. White House framework declares training on copyrighted material legal while UK maintains different standard: regulatory fragmentation extending across jurisdictions simultaneously.

Guardian documents FBI mass surveillance capabilities operating without AI: institutional surveillance infrastructure context for ongoing Anthropic-Pentagon confrontation over bulk data analysis.

White House national AI framework explicitly seeks to preempt state AI legislation, targeting four states with comprehensive laws and dozens with pending bills; child safety is the one domain where state authority is preserved; framework declares AI training on copyrighted material legal while Anthropic's $1.5B copyright class action settlement simultaneously seeks final court approval.

Essex Police pause facial recognition camera use after study finds racial bias: UK local enforcement withdrawing from deployment absent national regulatory framework.

BMG files copyright infringement lawsuit against Anthropic alleging Claude trained on pirated lyrics and reproduces copyrighted songs including "Uptown Funk" and "What a Wonderful World": major rights holder treats AI training on creative works as actionable infringement.

AI-generated public comments (20,000+ via CiviClick) influenced SoCal clean heat regulatory outcome with no agency framework to distinguish authentic from manufactured input: democratic participation mechanism vulnerability now documented with concrete regulatory consequence.

Virginia legislature passes balcony solar bill 96-0, becoming second state (after Utah) to classify plug-in solar up to 1,200W as consumer appliance requiring no utility permission or interconnection approval; 30 similar bills pending in other states.

UK Society of Authors launches human-authorship certification logo for books: private credentialing standard filling regulatory gap in absence of statutory AI content labeling requirements.

Ars Technica raises open-source license question for AI-rewritten code: whether AI systems can legally modify or relicense open-source software entering legal discourse without regulatory framework.

Guardian reveals UK government's multibillion-pound AI investment drive built substantially on phantom investments: announced commitments not matched by actual capital deployment.

Amodei publishes formal response to supply-chain risk designation arguing 10 USC 3252 applies narrowly to direct DoD contracts and requires least restrictive means; Anthropic simultaneously confirms Claude provided at nominal cost to active military operations to avoid disruption: legal challenge and operational continuity coexist.

Politico: lobbyists and ex-officials call Pentagon-Anthropic spat "bitterly ironic," say it undermines administration's broader AI agenda; Wired reveals Pentagon tested OpenAI models through Microsoft despite ban: two-track enforcement architecture documented.

Seven tech companies (Google, Meta, Microsoft, Oracle, OpenAI, Amazon, xAI) sign nonbinding White House "Ratepayer Protection Pledge" on data center electricity costs; Harvard Law energy expert calls it "theater," noting White House lacks enforcement authority over utility rate structures; residential electricity up 6% nationally, 16-19% in data-center-heavy states.

Virginia HB 434 establishes first state mandate for utilities to measure and optimize existing grid utilization: structural efficiency as regulatory category, distinct from capacity buildout mandates.

Supreme Court declines to hear Thaler AI copyright appeal, leaving intact lower court standard that human authorship is a bedrock requirement; AI-generated works enter public domain by default.

Atlantic sourcing details final Anthropic-Pentagon breakdown: bulk domestic surveillance demand (not autonomous weapons) was the terminal sticking point; OpenAI deal preserves EO 12333 compliance framing, which privacy advocates identify as the NSA's framework for capturing American communications offshore: two nominally identical red lines with structurally different enforcement architectures.

NRC publishes proposed technology-neutral regulatory framework for commercial fusion energy with 90-day comment period; Max Tegmark (MIT) publicly argues absence of binding AI regulation, product of industry lobbying, left companies without structural protection when state power demanded compliance.

Pentagon supply chain risk designation applies novel statutory authority (10 USC 3252) previously reserved for foreign-linked entities to a domestic AI company for maintaining safety policies; bipartisan senators urge ceasefire; framed by Politico as "attempted corporate murder" and potential partial nationalization signal.

Anthropic-Pentagon confrontation reaches public resolution: Pentagon threatened Defense Production Act invocation and supply chain risk designation; Amodei statement flags contradiction of simultaneously treating Anthropic as security threat and essential national security resource; AI policymakers warn approach could chill government-Silicon Valley partnerships.

27 U.S. states have introduced balcony solar legislation in 2026 legislative sessions; bipartisan coalitions (Missouri Republican, New York Democrat) driving bills to legalize plug-in panels without utility permission: state-level energy democratization advancing independent of federal action. Guardian: facial recognition error leads to wrongful arrest of Asian man for burglary 100 miles away: misidentification liability gap in live deployments without regulatory redress framework.

Dozen state legislatures advancing VPP bills (Michigan, Minnesota, New Jersey, Pennsylvania among active states); builds on laws already passed in Illinois, Virginia, Maryland, Colorado: state-level grid modernization advancing independent of federal action. Ireland makes basic income program for artists permanent: first national-level policy institutionalizing income floor in response to AI displacement of creative labor.

FDA proposes "plausible mechanism" approval pathway for rare-disease gene therapies tested in handful of patients; 60-day comment period; signals regulatory architecture bending to accommodate precision medicine that cannot meet traditional trial thresholds.

61 privacy regulators (ICO, EDPB, EDPS, Hong Kong, and others) issue coordinated Joint Statement on AI-generated imagery: frames non-consensual synthetic content as fundamental rights violation; invokes existing legal authority: contrast with New Delhi Declaration's voluntary approach.

New Delhi Declaration adopted by 88 countries: voluntary, non-binding, frames AI as shared human infrastructure around seven pillars; explicitly endorses open-source approaches. Dozens of countries declined safety commitments. OpenAI debated whether to alert police about Canadian shooter's AI chats: unresolved questions about platform liability and duty to warn.

13 state AGs sue DOE for terminating ~$8B in IRA/IIJA clean energy funding, alleging unconstitutional bypass of congressional appropriations.

EPA sued. NIST AI Agent Standards Initiative. Macron defending EU AI rules.

Illinois data center tax pause. FDA reversed Moderna refusal.

European Parliament blocks AI on lawmakers' devices.

EU AI Act Article 6 enforcement Aug 2026, 18 states failed transposition.

Pentagon threatening Anthropic over safety commitments. Military deployed Claude in Venezuela (83 casualties). Other labs agreed to drop guardrails.

DOE lost 85% clean energy staff ($27B at risk). TVA keeping coal plants.

Anthropic researcher public resignation warning. OpenAI researcher resigns over ads.

FDA refusing Moderna mRNA flu vaccine. California: 90-day AI layoff notice.

$35B clean energy canceled (policy uncertainty). EPA revoking endangerment finding.

NY data center moratorium proposed. Indiana: 12+ projects stopped.

19 states with privacy laws. Texas $1B+ settlement. Kentucky Character.AI enforcement.