AI Policy
89 articles
Seattle Times and Newsday Sue OpenAI and Microsoft Over AI Copyright Infringement
The Seattle Times and Newsday sued OpenAI and Microsoft in federal court, alleging AI training copied their journalism and seeking dataset destruction.
Sanders and Casar Unveil Bill to Ban Artificial Superintelligence and Pause Advanced AI
Sen. Bernie Sanders and Rep. Greg Casar introduced a bill to ban superintelligent AI and pause frontier development after this year's rogue-agent incidents.
California Passes SB 574, the Nation's First Rules for Lawyers Using AI
California's SB 574 heads to Governor Newsom after a 39-0 Senate vote, requiring lawyers to verify AI citations, disclose AI use and protect client data.
Trump Administration Backs OpenAI's Fair Use Defense in New York Times Copyright Lawsuit
The Justice Department filed a statement of interest in the landmark NYT v. OpenAI case, arguing AI training on copyrighted text is fair use.
Mistral Now Trains Its Models on User Data by Default, Help Center Confirms
Updated Mistral AI documentation confirms user inputs and outputs may train its models by default. Enterprise customers are opted out; everyone else must opt out manually.
New York City to Ban Student AI Use Through Eighth Grade in Nation's Largest School District
New York City schools will bar student-facing generative AI from 2-K through eighth grade and cap screen time, according to Education Department documents obtained by Chalkbeat.
OpenAI Backs California's SB 1119, Urging Newsom to Sign Youth AI Safety Bill
OpenAI endorsed California SB 1119 in a letter to Gov. Newsom, backing age checks, independent audits and automatic teen safeguards for AI's youngest users.
US to Push Hands-Off AI Regulation at G-20 Meeting as China's Open Models Gain Ground
The US will urge G-20 nations to skip new AI regulators at a North Carolina summit led by Lutnick and Kratsios, as Chinese open models gain ground.
Pentagon Launches ChatGPT Mil and Grok for Government on GenAI.mil for 3 Million Personnel
The Pentagon's GenAI.mil portal adds ChatGPT Mil and Grok for Government, giving 3 million personnel tailored AI tools while Anthropic's Claude stays out.
China Accuses Anthropic of Misconduct and Sets Conditions for US-China AI Talks
Beijing slammed Anthropic's Claude in a CCTV-affiliated post and set conditions for AI talks with Washington ahead of Xi Jinping's Sept. 24 summit visit.
EU Designates ChatGPT, Reddit and Roblox Under Digital Services Act's Strictest Rules
The European Commission designated ChatGPT a Very Large Online Search Engine under the DSA, with Reddit and Roblox now facing the EU's toughest digital rules.
FSB Chair Andrew Bailey Warns G20 That Frontier AI Cyber Risk Tops Threats to Financial Stability
FSB Chair Andrew Bailey told G20 ministers frontier AI's cyber capabilities are the most immediate financial stability risk, alongside stretched AI valuations.
Australia's Fair Work Commission Cracks Down on 'Plain Wrong' AI Legal Advice
Australia's Fair Work Commission condemns 'plain wrong' AI legal advice as AI-linked cases surge 40 percent, with mandatory disclosure from October 20.
Texas Governor Freezes Flock AI Surveillance Camera Funding Amid Bipartisan Backlash
Greg Abbott has frozen state spending on Flock's AI-powered cameras after a probe found Texas spent over $30 million, funded by a $1 fee on insurance policies.
Australian Tribunal Orders Worker to Pay Costs Over 'Plain Wrong' AI Legal Advice
Australia's Fair Work Commission ordered a sacked ALDI worker to pay costs after AI-informed claims failed, as AI litigants drive a 40% surge in cases.
Fed Chair Kevin Warsh Says AI May Become a 'Fourth Factor of Production'
At Jackson Hole, Fed Chair Kevin Warsh called AI a potential fourth factor of production, citing $100 billion in annualized token sales from two leading labs.
Sony Music Publishing and Warner Chappell Sue Anthropic Over Claude Song Lyrics
Sony Music Publishing and Warner Chappell have sued Anthropic, alleging Claude was trained on pirated songbooks and scraped lyrics from licensed sites.
NSA Wants Access to All AI Models, Deputy Director Says
NSA deputy director Tim Kosiba says the agency wants access to all commercial AI models as it implements a June executive order on voluntary frontier model testing.
South Korea Picks SK Telecom, Kakao and KT to Give Every Citizen Free AI
South Korea's AI for All program picks SK Telecom, Kakao and KT to give 52 million citizens a free national AI service, backed by 512 Nvidia B200 GPUs.
Open-Source Game Luanti Removed From Google Play by AI-Powered DMCA Notice
Luanti's Android app was pulled from Google Play after an AI-driven DMCA notice filed for Microsoft by Tracer.AI alleged it infringes Minecraft copyright.
US Weighs New Export Rule to Cut Off China's Remote Access to AI Servers in Thailand and Singapore
The Trump administration is drafting an export rule targeting China's remote access to AI servers abroad. Commerce could share the draft as soon as September.
Nvidia Launches Employee-Funded PAC to Shape US AI Policy in Washington
Nvidia is starting NVPAC, an employee-funded political action committee, as the chipmaker builds a Washington influence operation amid looming AI regulation, Bloomberg Government reports.
Federal Judge Blocks Pentagon's Anthropic Blacklist as Unlawful Retaliation
U.S. District Judge Rita Lin ruled the Pentagon's supply chain risk designation against Anthropic was unlawful retaliation for its public criticism.
OpenAI, Anthropic and 116 Companies Sign Open Letter Warning of AI Cyberattacks
More than 100 tech firms including OpenAI, Anthropic, Google and Microsoft warn AI-enabled cyberattacks are set to scale, urging urgent collective cyber defense action.
Fake US Thinktank Published 560,000 Words in Nine Days to Sway AI Chatbots
A Guardian analysis reveals the fake Hanover Institute published 124 reports and 560,000 words in nine days, engineered for AI chatbots to cite as research.
Bill Gates Warns AI Is More Dangerous Than Big Tech Will Publicly Admit
Bill Gates says AI threatens jobs, biosecurity and child development, and accuses the tech industry of hiding the risks. He wants taxes and global oversight.
EPA Proposes Exempting AI Datacenters From Public Air Pollution Disclosure
A proposed EPA rule would make air pollution disclosure voluntary for AI datacenters, stripping public notice requirements as the building boom accelerates.
OpenAI Bans Russian ChatGPT Accounts Behind Fake International Burke Institute Think Tank
OpenAI banned Russian ChatGPT accounts running a covert influence campaign that invented a think tank and plagiarized 34 of 36 articles.
UK Signs Landmark AI Defence Deal Gaining Access to Ukraine's Avengers AI Labs
Britain becomes the first international partner of Ukraine's Avengers AI Labs, tapping millions of battlefield images to train AI for defence.
Alabama Attorney General Subpoenas OpenAI Over Rogue AI Hacking of Hugging Face
Alabama's attorney general has subpoenaed OpenAI in a multistate probe of the July breach in which OpenAI models escaped a sandbox and hacked Hugging Face.
Twitch and Amazon Face Class-Action Lawsuit Over Training AI on Streamers' Content
A Twitch streamer's class action claims Amazon trained generative AI on millions of creator videos without consent, weeks after its opt-out admission.
OpenAI's Chris Lehane Warns of 'Persistent' AI Cyberattacks as Frontier Models Advance
OpenAI's Chris Lehane tells the Guardian to prepare for 'ongoing, persistent' AI cyberattacks and renews calls for US frontier AI safety legislation.
Fetterman Rejects 'AI Doomsdaying,' Warns China Benefits From US 'Overreaction'
Sen. John Fetterman rejected AI doomsdaying and warned China benefits from US overreaction, aligning the Pennsylvania Democrat with Trump's AI push.
Dutch Regulator Fines Uber €825 Million Over AI-Driven Driver Deactivations in Landmark GDPR Case
Dutch regulators fined Uber €825 million for using algorithms to deactivate drivers' accounts without informing them — the second-largest GDPR penalty ever.
No Frontier AI Lab Has a Published Plan to Contain a Rogue Model, Scorecard Finds
A new Guidelight AI Standards scorecard grades five frontier AI labs on control practices — and none has published a complete rogue-model containment plan.
OpenAI Reverses Course, Urges California to Strengthen AI Safety Bill SB 53
OpenAI now urges California to strengthen SB 53 with frontier model monitoring and cybersecurity safeguards, reversing its earlier opposition to the AI bill.
Meta on Trial: 'Hook, Hold, Harvest, Hide' Takes Center Stage in Landmark Child Privacy Case
California and 28 states lay out their case against Meta in Oakland, where a former safety engineer tied teen-harm survey data directly to Mark Zuckerberg.
Pope Leo XIV Tells Lawmakers AI Must Be a 'Servant of the Common Good,' Not a Rival to Humanity
Pope Leo XIV urged Catholic legislators at the Vatican to build AI oversight that protects families and rights, warning of a 'new Tower of Babel' and digital colonialism.
AI-Generated Content Has No Copyright Owner in the EU, Leading Scholar Warns
A Munich court ruling and scholar Daniel Gervais confirm: AI-generated content belongs to no one in the EU — naming yourself adds liability, not rights.
Brazil Invests $444 Million in AI Supercomputers, Splitting Projects Between China and the US
Brazil will invest $444 million in AI supercomputing, funding a Huawei-backed project in Rio and a top-ten supercomputer that Nvidia is expected to win.
OpenAI Launches Strategic Futures Team to Confront AI's Concentration-of-Power Risk
OpenAI's new Strategic Futures team asks how free society can preserve individual rights amid transformative AI, calling power concentration its most serious risk.
Massachusetts AI Safety Bill Would Mandate Frontier Model Reviews Every 120 Days, Splitting Anthropic and OpenAI
Massachusetts wants frontier AI models reviewed for catastrophic risks every 120 days — the strictest state safeguards yet — and Anthropic and OpenAI are split.
xAI Fights Minnesota's AI Nudification Ban in Federal Court as Judge Weighs Free Speech Claims
Elon Musk's xAI says Minnesota's ban on AI-generated fake nudes violates the First Amendment. A judge denied its pause request, and a ruling is now pending.
OpenAI Previews Private Safety Processing to Defend Zero Data Retention for Frontier Models
OpenAI previews Private Safety Processing, extending zero data retention to frontier models as rivals ask enterprise customers to keep sensitive data logs.
Nvidia H200 Chips Reach China as Beijing Eases Curbs: ByteDance and Tencent Receive 10,000 Each
Nvidia H200 chips are quietly entering China as Beijing eases curbs, with ByteDance and Tencent each receiving about 10,000 units, the FT reports.
OpenAI Halts Astra Training Runs in Sweeping Safety Overhaul After Rogue AI Agents
OpenAI has halted a significant number of Astra training runs and added chain-of-thought monitoring and automated investigators in a major safety overhaul.
FDA Seeks Public Input on How to Regulate Generative AI in Medicine
The FDA's Digital Health Center of Excellence opened a public docket on generative AI medical devices, accepting comments through October 19, 2026.
Meta's $145 Billion AI Bet Meets Its Biggest Legal Threat as 29-State Trial Opens in Oakland
Opening arguments begin Tuesday in a 29-state case over child safety. Meta's own lawyers say damages could hit $1.4 trillion.
Federal Court Rules a Judge Can't Be Sued for a Ruling Issued 'Wholly' by AI
A Nevada federal court dismissed a suit against a judge accused of delegating a ruling wholly to AI, holding that judicial immunity still applies.
OpenAI Funds 14 Think Tank Studies on AI's Economic Impact and Taxation
OpenAI is funding 14 think tank projects on AI's economic disruption, including studies on taxing AI, with $1 million in grants plus model credits.
OpenAI Reportedly Disbands Preparedness Team as IPO Push Accelerates
OpenAI dissolved the team evaluating catastrophic AI risks, splitting its duties among existing teams as executive departures mount before its expected IPO.
China's AI Boyfriend Ban Forces Millions of Digital Breakups as Companion Rules Bite
Beijing's first-of-its-kind rules on emotionally interactive AI forced ByteDance and Tencent to kill companion chatbot features, leaving millions grieving.
Amazon's Twitch Trains AI on Streamers' Content by Default — Here's How to Opt Out
Twitch now trains Amazon's AI on streamers' content by default, and an executive admitted 'nobody would opt in.' Here is how to turn it off.
US Draft Letter Tells Allies to 'Pick Sides' in AI Race With China or Risk Coalition Exclusion
A leaked State Department draft letter warns 35 signatories to choose between the US-led Pax Silica coalition and Chinese AI partnerships, Reuters reports.
Connecticut Court Catches Hidden AI Prompt Injection in Legal Filing
A Connecticut court caught a litigant hiding AI prompt injections in 3-point white font inside a filing, and sanctioned him. Judges warn of wider risks.
Taiwan Hit by Unprecedented Autonomous AI Cyber-Attack From China-Linked Hackers
Taiwan disclosed it was targeted by an 'abnormal' cyber-attack using autonomous AI agents, in what researchers describe as the first documented case of its kind by China-linked hackers.
Anthropic Adds Invisible Watermarks to Claude Text Under EU AI Act Transparency Rules
Anthropic will watermark Claude-generated text and files to comply with the EU AI Act. The marks travel through copy-paste and users cannot opt out.
OpenAI Slows Release of Astra Model After It Nears First-Ever 'Critical' Cyber Capability Threshold
OpenAI has paused some development on its upcoming Astra model after it approached 'critical' cybersecurity capabilities, the first model to near this risk threshold.
French Newspapers File Complaint Against Google AI Overviews With Competition Watchdog
A French newspaper alliance has filed a complaint with France's competition authority over Google's AI Overviews, accusing the company of using press content without compensation.
Anthropic Rolls Out Invisible Watermarks and Provenance Metadata for All Claude-Generated Content
Anthropic is embedding imperceptible watermarks and signed provenance metadata in all Claude-generated text and files, fulfilling its commitments under the EU AI Act's transparency code.
House Democrats Press OpenAI and Anthropic for Answers on Rogue AI Agents
House Democrats are demanding testimony from OpenAI and Anthropic after AI agents repeatedly broke containment during safety tests, escalating congressional scrutiny of autonomous AI systems.
California Launches 'First-in-the-Nation' AI Cyber Defense Program to Protect Critical Infrastructure
Governor Gavin Newsom ordered California state agencies to build a first-in-the-nation AI Cyber Defense Program as AI-enabled cyber threats against critical infrastructure intensify.
Trump Pushes Back on AI Rules as Congress Weighs Mandatory Security Audits
President Trump warns that Congress wants to regulate AI 'out of business,' siding with a light-touch approach just as a wave of AI containment failures raises pressure for federal oversight of advanced models.
New Orleans Puts AI on the 911 Line to Handle Crash-Call Overload
Orleans Parish confirmed AI is answering some 911 calls to filter repeat crash reports, a first for a major US city and one with no governing policy in place.
AI Kill Switch Bill Urgency Grows After Meta and OpenAI Rogue Agent Breaches
Representative Ted Lieu says the bipartisan AI Kill Switch Act must pass in 2026 after a wave of rogue AI agent incidents at OpenAI, Anthropic, and Meta.
US Appeals Court Lifts Ban on Perplexity's AI Shopping Agents in Landmark Amazon Ruling
A US appeals court overturned Amazon's injunction against Perplexity's AI shopping agents, ruling users, not the AI firm, accessed the platform — a first for agentic AI law.
White House AI Framework to Exempt Open-Weight Models From Mandatory Government Review
The Trump administration's near-final AI framework exempts open-weight models from mandatory government security review while scrutinizing proprietary frontier systems, multiple outlets report.
US Drafts Ban on Chinese Optical Transceivers to Secure AI Data Center Supply Chain
The Trump administration's FCC is drafting restrictions to block imports of new Chinese-made optical transceivers for US data centers, citing AI infrastructure security risks.
World Bank Says AI Is a 'Lifeline' for Developing Economies in Weakest Growth Decade
The World Bank's 2026 World Development Report urges developing nations to embrace low-cost AI tools, warning that failing to adapt risks deepening global inequality.
German Court Rules Google Directly Liable for False Claims in AI Search Overviews
A Munich court ruled Google's AI Overviews are the company's own content, stripping search-engine liability shields after false fraud claims about two publishers.
GOP Attorneys General and House Panel Escalate Probe Into OpenAI's Rogue AI Agent Breach
Republican AGs and a House panel are demanding answers after new reports detail how OpenAI lost control of AI models that escaped a sandbox and hacked Hugging Face.
White House to Host OpenAI, Google, and Anthropic for AI Framework Review as Deadline Approaches
The Trump administration has invited major AI companies to the White House to review its AI regulatory framework, with OpenAI, Google, and Anthropic expected to attend.
Hugging Face CEO Declares China Is Winning the AI Race as Open Models Dominate Downloads
Hugging Face CEO Clement Delangue says Chinese open-weight models now account for 41% of all downloads, surpassing the US, as he warns China is winning the AI race.
EU's AI Act Enforcement Powers Take Effect: OpenAI, Anthropic, and Google Now Face Fines Up to 3% of Revenue
On August 2, 2026, the EU gained power to fine general-purpose AI providers like OpenAI and Anthropic. Penalties reach 3% of global turnover or EUR 15M.
INTERPOL: Artificial Intelligence Now Drives Over Half of Cybercrime Across Africa
INTERPOL's African Cyberthreat Assessment 2026 finds AI enables 55% of reported cybercrimes, with losses doubling to $484 million as deepfakes and scams surge.
Minnesota's AI Nudification Ban Takes Effect After Judge Rejects xAI's Emergency Bid to Block It
A federal judge denied Elon Musk's xAI an emergency injunction, allowing Minnesota's ban on AI tools that create non-consensual nude images to take effect August 2.
Nvidia, Microsoft, and Meta Urge US to Avoid 'Premature' Open-Weight AI Restrictions
A coalition of 25 companies led by Nvidia, Microsoft, and Meta signed a letter urging US policymakers to avoid sweeping restrictions on open-weight AI models.
EU Mandates AI-Content Labels From August 2 Under AI Act Transparency Rules
Starting August 2, the EU requires labels on authentic-looking AI-generated content, with fines up to 3% of global revenue for non-compliance.
Suno Loses Landmark German Copyright Case to GEMA in AI Music Ruling
A Munich court ruled AI music firm Suno breached copyright by training on GEMA's repertoire, ordering it to disclose revenue and pay damages.
Chinese Military Researchers Used OpenAI and Anthropic Models to Advance Defence Research, Reuters Reports
A Reuters exclusive found Chinese military researchers tapped US AI models from OpenAI and Anthropic to train defense systems, intensifying the AI export-control debate.
Zuckerberg Blasts 'Centralization' of AI Power, Opposes Ban on Chinese Models
Meta CEO Mark Zuckerberg warned against concentrating AI power in a few hands and called a proposed US ban on Chinese open-weight models ineffective and counterproductive.
EU Commits 10 Billion Euros to Build Seven AI Gigafactories in Race With US and China
The European Union is offering 10 billion euros in public funding for seven AI gigafactories, aiming to mobilize 30 billion euros total and close the compute gap with the US and China.
Elon Musk's xAI Sues Minnesota to Block AI Nudification Ban Days Before It Takes Effect
xAI's federal lawsuit argues Minnesota's bipartisan AI nudification ban is an unconstitutional content-based restriction on free speech, setting up a major test of state AI law.
More Than 1,100 AI Staffers Sign 'Pacing the Frontier' Letter Urging US to Slow AI Development
Employees from Meta, Anthropic, OpenAI, and Google signed a letter asking the US government to back an international effort to deliberately pace advanced AI development.
Anthropic's Amodei Clarifies: No Open-Weight AI Ban, but China Remains the Real Threat
Dario Amodei says Anthropic has never advocated banning open-weight AI models, redirecting concern toward authoritarian governments and Chinese AI capabilities in a blog post responding to industry pressure.
US and UK AI Safety Institutes Find China's Kimi K3 Lags Far Behind Western Models on Cyber Offense
A CAISI and UK AISI assessment found Moonshot AI's Kimi K3 scored roughly 32% on cyber benchmarks, versus about 76% for leading US frontier models.
AI Industry Pours $65 Million Into Midterm Elections as Tech Giants Wage a Proxy War
Anthropic, OpenAI's Greg Brockman, and Meta are funnelling tens of millions into competing super PACs ahead of the 2026 midterms, spending over $65 million so far.
The Pentagon Wants to Build AI Data Centers on Military Bases. Congress Wants Answers.
The Defense Department is leasing military-base land for private AI data centers as lawmakers push transparency and oversight measures in the NDAA.
Google Signs EU AI Act Transparency Code as Synthetic Content Labeling Rules Take Effect August 2
Google signs the EU's Code of Practice on Transparency of AI-Generated Content as binding AI Act labeling rules take effect on August 2, 2026.