Policy & Regulation
Trump Signs Executive Order Establishing Voluntary Pre-Release Review for Frontier AI Models. The order asks AI companies to voluntarily submit their most powerful models for government testing up to 30 days before public release, while explicitly rejecting mandatory licensing. It revokes Biden-era frameworks and shifts federal AI policy toward a cybersecurity-focused, deregulatory posture.
White House Issues National Security Memo on AI in Defense and Intelligence. NSPM-11 sets governance boundaries for AI use in national-security systems, including surveillance limits and autonomy in weapons. It ties AI safety and civil-liberties concerns directly to defense and intelligence policy.
Rep. Sara Jacobs Introduces Bill to Hold AI Systems Legally Accountable. The proposed legislation would create legal liability when AI systems cause unlawful harms in employment, housing, lending, and healthcare. It reflects a growing congressional push to close the accountability gap for automated decision-making.
EU Reaches Deal on First Amendments to the AI Act. The “Digital Omnibus” provisional agreement pushes back compliance deadlines for high-risk AI systems to late 2027 and introduces a strict new ban on AI-generated non-consensual intimate imagery and child sexual abuse material. Consumer advocates warn the changes “roll back key protections” and create loopholes.
State-Level AI Bills Are Moving Fast: Weekly Legislative Roundup. Vermont banned therapy bots, Illinois sent five AI bills to the governor (including a frontier model safety act and a ban on AI-enabled rental price fixing), and California advanced 30 AI bills in committee. New York and Rhode Island are also finally seeing movement.
OpenAI Publishes Blueprint for Federal AI Governance. OpenAI proposes a U.S. federal framework for frontier AI safety, building on state laws like California’s SB 53 and strengthening the Center for AI Standards and Innovation. The document calls for durable institutions that can evolve alongside the technology.
Illinois Postpones Rules on AI-in-Hiring Transparency Requirements. The state’s Department of Human Rights delayed proposed rules implementing AI-in-employment notice requirements under Public Act 103-0804. It is a concrete example of how labor-market AI regulation is hitting friction in the implementation details.
The Data Center Wars
California City Votes 86% to Permanently Ban Data Centers. Monterey Park, California became the first U.S. city to pass a ballot initiative permanently banning construction of power-hungry AI data centers. The vote reflects growing public frustration over environmental, public health, and utility rate impacts of AI infrastructure.
Seattle Poised to Become Largest U.S. City to Ban New Data Centers. A proposed moratorium in Amazon and Microsoft’s backyard would make Seattle the biggest American city to block new data center construction. The measure highlights how local government pushback over energy and water demands is reaching the tech industry’s home turf.
Police Are Surveilling Americans Who Criticize AI Data Centers on Social Media. A confidential law enforcement bulletin obtained by The Intercept reveals that people speaking out against AI data centers are being tracked as potential threats. A civil rights lawyer said he was troubled by the fusion center’s association of AI skeptics with terrorists over “legitimate, popular political concerns.”
Environmental Lawsuits Are Becoming a Real Obstacle for the Data Center Boom. Communities and environmental groups are turning to litigation to block massive AI data center projects, with recent legal wins including a Minnesota court order against a Google-backed project and an NAACP injunction targeting unpermitted gas turbines at xAI’s Memphis facility.
Think Tank Offers Policy Framework for Communities Facing the Data Center Rush. The Institute for Local Self-Reliance argues that tech firms projected to spend $700 billion in capital expenditures in 2026 are shifting costs onto local communities. The report offers a framework to reassert public authority and curb monopoly power in infrastructure siting.
Ethics, Safety & Deepfakes
UK Labour MP Files Test Case Against xAI Over Grok Deepfake Images. MP Jess Asato’s lawsuit in the UK High Court alleges that xAI’s Grok chatbot generated non-consensual sexualized deepfake images of her, seeking to establish platform liability for AI system design. SpaceX has earmarked over $500 million in its IPO filing for potential litigation losses linked to Grok.
CNN Sues Perplexity AI Over Unlawful Distribution of News Content. The filing marks CNN’s first lawsuit against an AI company and possibly the first from a broadcaster. The New York Times, New York Post, Dow Jones, and Reddit have all filed similar suits against the startup. Perplexity’s response: “You can’t copyright facts.”
arXiv Announces One-Year Bans for Authors Who Upload “AI Slop”. Thomas Dietterich, chair of arXiv’s computer science section, announced a strict new policy targeting papers with clear evidence of AI-generated filler and hallucinated citations. The move reflects growing alarm in the scientific community about the erosion of research integrity.
Anthropic Report Warns That AI Systems Are Getting Better at Improving Themselves. The report explores rapid progress in AI systems writing their own code and designing successor models. The authors caution that this recursive self-improvement could outpace institutional readiness, raising fundamental questions about maintaining human oversight.
Economics & Employment
ILO Finds GenAI Is Reshaping Work Without Mass Displacement, But Deepening Inequality. This International Labour Organization brief synthesizes evidence from multiple countries on how generative AI is changing tasks, employment, and workplace dynamics. Large-scale job losses remain limited so far, but the technology is eroding opportunities for younger workers and altering worker autonomy.
New Paper Maps How Firms Are Reorganizing Hiring Around Generative AI. This SSRN paper studies how companies are changing hiring patterns and job content as generative AI spreads through the U.S. economy. It moves beyond simple displacement narratives to analyze how labor demand is being restructured across tasks, occupations, and locations.
Brookings: AI-Exposed Jobs Are Concentrated in Democratic-Leaning Counties. Analysis reveals that 62 of the top 100 U.S. counties most exposed to AI-driven job disruption voted blue in 2024. The finding highlights potential flashpoints for economic anxiety and partisan policy debates heading into election season.
China Is Mobilizing Massive Human Labor to Build Its Robotics Edge. Rest of World reports on how China is organizing large-scale human data collection to train “physical AI” and robotics systems. The piece is strong on how AI can reorganize labor rather than simply replace it, with global competitive implications.
Research
Nature: Can Social Science Survive AI-Contaminated Survey Data?. LLMs are polluting social science surveys with AI-generated responses (up to 45% in some tests), generating spurious findings and undermining public trust in research. The article also explores how AI tools could help detect errors and handle complex data if deployed carefully.
New Framework Aims to Cut Through Polarized AI Hype and Doom Narratives. This arXiv paper introduces the VET (Valence, Effectiveness, Trajectory) framework for analyzing how AI is discussed in media and public debate. It is designed as an AI literacy tool for policymakers and the public to counter exaggeration and improve understanding of what AI can and cannot do.
Last Updated: 2026-06-07 09:34 (California Time)