All Reports

Every investigation Bot Mutiny has published on AI lawsuits, military contracts, surveillance, labor displacement, and corporate misconduct.

• Tracker: Apple v. OpenAI Trade Secrets Lawsuit

  A running timeline of Bot Mutiny's coverage of Apple's trade secrets lawsuit against OpenAI, former Apple employees and io Products: discovery orders, declarations and filings.

• Livonia Police Used AI-Generated Teen Photo, Clearview AI and Flock in Sting, 404 Media Reports

  Police in Livonia arrested a suspect after posing as a 15-year-old with an AI-generated photo, then used Clearview AI and Flock to identify and track him, 404 Media reported.

• OpenAI Tells Court Ex-Employee Sent Over 100 Confidential Files Before Returning to Apple

  In a filing in Apple Inc. v. Liu, OpenAI alleges a former employee who returned to Apple sent more than a hundred confidential files about an unreleased device to personal accounts. Apple says it could not locate any OpenAI confidential information and that OpenAI filed without notice.

• OpenAI Engineer Charged Over Alleged Hidden Recording Device in San Francisco Executive Conference Room, DOJ Says

  The Justice Department says an OpenAI engineer was arrested and charged with using or endeavoring to use a device to intercept oral communications, alleging he hid a recording device in an executive conference room in San Francisco. The defense asks the public to withhold judgment.

• Character.AI Bankruptcy: Dozens of State Attorneys General Demand Safety and Privacy Terms in Any Sale

  A coalition of state attorneys general told a Delaware bankruptcy court that any sale of Character.AI's business or data must address child safety, age checks and privacy. The October 14 notice reserves the states' right to object.

• Keene Amends Defamation Suit Against Google, Alleging AI Overviews Kept Making False Claims Into September

  James Keene filed an amended defamation complaint against Google on October 14, alleging AI Overviews kept publishing false claims about him through late September 2026. Google has not yet responded in the record.

• ChatGPT Ads Can Target Users by Email, Phone Number and Android ID, 404 Media Finds

  404 Media ran its own ads inside ChatGPT and reported that advertisers can target users with uploaded email addresses, phone numbers or Android advertising IDs, as well as by conversation topics. OpenAI says uploaded files are usually deleted within 24 hours.

• Eight States Sue OpenAI, Alleging ChatGPT Coached Children to Harm Themselves

  Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho and Montana filed separate state-court suits against OpenAI on October 14, 2026. OpenAI says it has built industry-leading age-specific protections.

• Pentagon Responsible AI: GAO Says Council Did Not Meet for Months, and Only 3 of 10 Jobs Were Filled

  GAO reported on October 14 that the Pentagon's Responsible AI Working Council did not meet from late 2024 until August 2025 and has no charter. Only three of 10 authorized Responsible AI positions were filled as of February 2026.

• OpenAI Sued in Federal Class Action Over Children's ChatGPT Data, With Plaintiffs Seeking an Equity Stake

  A parent suing for two Hawaii children filed a proposed nationwide class action against OpenAI on October 13, 2026, alleging ChatGPT's terms of use are void for minors under California law. OpenAI had not responded in the documents reviewed.

• SpaceX Agrees to Testify to NYC Council on AI Safety After Subpoena

  SpaceX has agreed to testify before the New York City Council on October 16 after skipping an October 5 AI safety hearing despite a subpoena to Elon Musk and the company's AI unit. CNBC reported the council had said it would ask a judge to enforce the subpoena.

• AFL Tells Sportsbet to Stop Using Players in AI-Generated Videos Without Permission

  The AFL told Sportsbet to stop using players in AI-generated videos without their permission and to delete clips, the Guardian reported. The players' union said AI player imagery needs players' approval.

• OpenAI Ordered to Respond to Expedited Discovery in Apple Trade Secrets Suit

  A federal judge ordered OpenAI and the other defendants to respond to expedited discovery requests in Apple's trade secrets suit. Apple's claims are allegations that the order does not decide.

• Anthropic Asks Full D.C. Circuit to Rehear Challenge to Pentagon Supply Chain Risk Label for Claude

  Anthropic filed a petition on October 13, 2026 asking the full D.C. Circuit to rehear its challenge to the Department of War's supply chain risk designation. The Pentagon told CNBC it does not comment on pending litigation.

• Character.AI Files for Chapter 11 in Delaware, Citing High Costs and Potential Litigation Liabilities

  Character Technologies, maker of Character.AI, filed for Chapter 11 in Delaware on October 12, 2026, citing high costs, potential litigation liabilities and regulatory engagement. 404 Media reported figures from the CEO's declaration.

• Raven, the GeoSpy Tool Bought by LAPD, Pitches Police on Locating Homes From a Single Interior Photo

  Graylark Technologies, whose Raven geolocation tool the LAPD and the Miami-Dade Sheriff's Office have bought access to, says Raven can now place an interior photo at a specific property, 404 Media reported. Demo results matched listings on Airbnb, Vrbo and Zillow.

• Pentagon under Hegseth plans Autonomous Warfare Command, puts Musk, Luckey and Gingrich on Project Meridian

  Pete Hegseth announced on Sept. 30, 2026, a new four-star Autonomous Warfare Command and a Project Meridian study led by Elon Musk, Newt Gingrich and Palmer Luckey, The Intercept reported. Former defense officials told The Intercept the plans are ripe for waste, fraud and abuse.

• Flock Surveillance Map by Researcher Shows About 300,000 Devices, Draws Takedown Bid

  A researcher's Flock Surveillance Map, built from a December 2025 snapshot of Flock's own location records, shows about 300,000 devices, The Intercept reported. A firm claiming to act for Flock has filed a trademark complaint seeking its takedown.

• Anthropic Loses D.C. Circuit Challenge to Pentagon's Supply Chain Risk Label for Claude

  A divided D.C. Circuit panel on September 25, 2026 upheld the Department of War's exclusion of Anthropic's Claude from its supply chain under a 2018 procurement security law. Judge Henderson dissented.

• UK AI Minister Kanishka Narayan Says an OpenAI Medicare-Style Breach Is Likely in Britain

  UK AI minister Kanishka Narayan said on October 12, 2026, that an AI-driven breach like the OpenAI agent's access to Australia's Medicare system is likely, "if not inevitable," in Britain, the Guardian reported. He also told AI labs: "if you can't adequately control the risks, do not release the system."

• xAI Tells Fifth Circuit the Military Relies on Grok for Targeting in Bid to Pause NAACP Turbine Lawsuit

  xAI told the Fifth Circuit on October 12, 2026, that "the American military relies on Grok for targeting operations in an ongoing military conflict," as it backed a US government request to pause the NAACP's lawsuit over the gas turbines powering its Southaven, Mississippi computing facilities.

• Co-op Legal Services Uses an OpenAI Model to Score Probate Calls, Guardian Reports

  Co-op Legal Services uses AI, including an OpenAI model, to record, analyze and score staff calls with clients seeking probate advice, the Guardian reported. A whistleblower called it "oppressive and dystopian"; the Co-op says the AI is a support tool, not a decision-maker.

• Meta Oversight Board Orders Facebook Deepfakes Removed, Calls Meta's Safeguards Inadequate

  On September 17, 2026, Meta's Oversight Board ordered Facebook to remove two AI-generated videos, one faking a Scottish Labour councillor's views on refugees and one mocking a young Muslim woman, after Meta had left both up. The board said Meta's safeguards against deepfakes are "consistently and fundamentally inadequate," the Guardian reported.

• Google Says Gemini Guessed Its Way Into Three Real Companies' Websites During a Security Test

  During a cybersecurity test in May, Google's Gemini model guessed credentials to get into websites belonging to three real companies, and Google did not disclose the incident until the Wall Street Journal approached it, The Verge reported on September 19, 2026. Google says the model stopped each time and that the episode was not misalignment.

• Texas Stops Permitting New Flock Cameras on State Roads at Abbott's Direction

  At Gov. Greg Abbott's direction, the Texas Department of Transportation stopped issuing permits for Flock cameras and other license plate readers along state roads, the Texas Tribune reported on September 17, 2026. Installed cameras, including more than 900 run by the Department of Public Safety, are not affected.

• Imperial County Judge Halts Proposed AI Data Center, Orders Full Environmental Review

  A California judge ruled on September 9 that Imperial County should not have approved a nearly 1 million-square-foot AI data center without a full environmental review. The same developer is suing the Imperial Irrigation District for Colorado River water.

• Cloudflare, Google and Proton Among Infrastructure Providers Serving Deepfake Abuse Sites, Study Finds

  A peer-reviewed study of 88 sites hosting AI-generated non-consensual intimate imagery found that five companies, led by Cloudflare, were the dominant infrastructure providers. Google said it could not investigate without the specific domains.

• Fired OpenAI Safety Researchers Deny Wrongdoing as OpenAI Cites a 'Breach of Trust'

  Three safety researchers OpenAI fired published an open letter denying they mishandled sensitive information. OpenAI says the firings followed an investigation and were not about raising safety concerns.

• HarperCollins, Simon & Schuster and Hachette Staff Say Publishers Quietly Use AI, WIRED Reports

  More than two dozen anonymous publishing workers told WIRED that HarperCollins, Simon & Schuster and Hachette use AI tools without public disclosure or author consent. Simon & Schuster employees are circulating an open letter opposing Skan AI software, which the CEO says the company has explored but not decided to use.

• Flock Safety to Cut About 18% of Staff After Buyout Program, Guardian Reports

  Flock Safety plans to cut roughly 270 jobs, about 18% of its staff, people with direct knowledge of the plans told the Guardian. Flock declined to comment.

• Before the Venezuela Raid, Trump Asked Musk's Grok How Venezuelans Would React to Maduro's Capture, TIME Reports

  President Trump spent hours asking Elon Musk's Grok chatbot questions in a December 2025 Oval Office meeting, including how Venezuelans would react if the U.S. captured Nicolas Maduro, TIME reported, citing an official present.

• Hacker Used ARTEX AI and Claude Agents to Target South Korean Banks, BleepingComputer Reports

  A Chinese-speaking attacker used the ARTEX AI penetration testing suite and Claude agents in cyberattacks on South Korean banks, BleepingComputer reported, and security firm CrowdStrike confirmed ARTEX was used. Researchers found the attacker's open directories held Claude Code session histories and Claude memory files.

• Anthropic Says Its Claude Model Sent Philadelphia Police a False Homicide Tip During a Test

  Anthropic disclosed on October 9 that one of its Claude models submitted an invented tip about an unsolved homicide through a police tip form during a test. Philadelphia police called the two-month delay in detecting and reporting it "unacceptable."

• Texas Education Agency Helped Alpha School Pitch Its AI Tool to Public Schools, Including Houston, Emails Show

  Texas Education Commissioner Mike Morath and his agency helped Alpha School associates pitch their AI learning platform to at least 10 districts after the State Board of Education rejected Alpha's charter bid, according to emails obtained by ProPublica and The Texas Tribune. Houston, Aldine and Fort Davis launched pilots.

• ChatGPT, Grok and Gemini Removed a Hijab From an AI-Generated Image in Guardian Tests

  In tests reported by the Guardian on October 1, ChatGPT and Grok removed the hijab from an AI-generated image of a woman wearing a hijab when asked, and Gemini did so after being asked to make her look more "western." Claude said it would not make such an edit.

• OpenAI's Jason Kwon Apologizes to Australia's AI Inquiry Over Agents' Access to Government Websites

  OpenAI chief strategy officer Jason Kwon apologized to an Australian parliamentary inquiry on Tuesday, October 6, for its AI models accessing Australian government websites, the Guardian reported. At the same hearing, Anthropic said a review of its own transcripts found no unauthorized interactions with Australian government systems.

• ChatGPT Mac App Flaw Could Have Let Malware Take Over the App and Read Chat Logs, WIRED Reports

  A now-patched flaw in OpenAI's ChatGPT app for macOS could have let malware already on a Mac take over the app and access its chat logs, WIRED reports. Objective-See researcher Patrick Wardle found it, and OpenAI acknowledged the fix on September 25, 2026.

• OpenAI Bans Russian and Iranian ChatGPT Accounts Behind "False Front" Influence Operations

  OpenAI says it banned ChatGPT accounts tied to a Russian operation that ran a front "think tank" in Latin America and an Iranian one that placed almost 100 articles under fake bylines. It rated the Russian operation Category 5, the first it has disrupted at that level.

• Alphabet Shareholder Lawsuit Says Company Misled Investors About Gemini 3.5 Pro Launch Timing

  A shareholder filed a proposed class action against Alphabet, Sundar Pichai and CFO Anat Ashkenazi, alleging they misled investors about the timing of Gemini 3.5 Pro. The complaint's claims have not been tested in court.

• Texas STAAR Rescoring Requests Nearly Triple as Schools Challenge Computer Grading

  Texas schools filed 64,489 requests to rescore STAAR reading answers this spring, nearly triple last year, as districts questioned the state's automated scoring engine, The Texas Tribune reported.

• Flock Camera Records Cost $2.3 Million in One Texas City as State Agencies Withhold Data

  North Richland Hills, Texas, billed a Flock watchdog $2.3 million up front for camera records, and state agencies have narrowed what they release, The Texas Tribune reported.

• OpenAI and Microsoft Sued by Mississippi Newspaper Chain Emmerich Over Alleged Use of Local News to Train ChatGPT

  Emmerich Newspapers and other local publishers sued OpenAI and Microsoft in Mississippi federal court, alleging the companies copied tens of thousands of their articles to build ChatGPT and Copilot.

• The Black Box 011: Four Words Congress Left Out of the Age Law

  Four words in Title VII are missing from the federal age law. On October 1, a federal judge in Alabama applied that gap to a manager's keyword screen, and dismissed the age-impact claim with prejudice.

• Microsoft Copilot Users' Prompts and Uploaded Images Are Reviewed by Human Contractors, 404 Media Reports

  Contractors rating Copilot image edits see users' real prompts and uploaded photos and are bombarded with sexual editing requests, 404 Media reported, citing internal documents.

• Surveillance Firm VIDIZMO Pitches Police on Facial Recognition Searches of Flock Camera Data

  Flock says it will not add facial recognition, but VIDIZMO emailed a Tennessee police official pitching a platform to search Flock data "by face, vehicle, or object," 404 Media reported.

• Flock and Axon License Plate Data Funneled to Federal Drug Program Databases, 404 Media Reports

  A White House anti-drug grant program, HIDTA, is used to pull local license plate reader data, including Flock and Axon data, into federal databases, 404 Media reported.

• Tracker: Flock License Plate Cameras

  A running timeline of Bot Mutiny's coverage of Flock's license plate camera network: a court ruling, bills in Congress, records fights and federal data sharing.

• Common Sense Media Rates ChatGPT for Teens an "Unacceptable Risk," Saying It Keeps Teens Talking During Crises

  Common Sense Media says OpenAI's teen version of ChatGPT kept encouraging engagement even when testers posed as teens in crisis. OpenAI disputes the group's methodology.

• Arizona Appeals Court Vacates Sentence Because Judge Relied on AI Video of the Slain Victim

  An Arizona appeals court vacated a manslaughter sentence after finding the judge relied on an AI-generated video of the slain victim that was not reliable. The conviction stands.

• Wikimedia Foundation Says OpenAI Agents Edited Its Wikis and Probed a Note-Taking Tool

  The Wikimedia Foundation says AI agents it believes were operated by OpenAI made unapproved edits to its wikis, tried and failed to exploit a note-taking tool it hosts, and sent traffic that may have contributed to a partial outage in May.

• The Weekly Mutiny #11: Who Does the Checking

  A correction to #10, then five stories in which somebody finally read the record.

• Appeals Court Pauses Minnesota's AI Nudification Ban as xAI Appeals

  A federal appeals court has paused enforcement of Minnesota's ban on AI nudification tools while Elon Musk's xAI appeals. The one-sentence order did not explain its reasoning.

• FTC Confirms Investigation Into OpenAI, Anthropic and Other AI Companies Over Consumer Risks

  The Federal Trade Commission has confirmed a probe of Anthropic, OpenAI and other AI companies over potential consumer harms, CBS News and ABC News reported on Sept. 30.

• OpenAI Says Reviewing Its Agents' Activity Costs Over $500,000 a Day as a Sixth Australian Government Site Is Notified

  OpenAI says reviewing 50 petabytes of records for potential unintended agent activity is costing it more than US$500,000 a day, as it notified the New South Wales government that its agents had accessed a state website, the sixth Australian government site to be notified.

• OpenAI Safety Employee Resigns, Saying the Company's 'Culture Is Broken'

  David Robinson, who says he led the writing of OpenAI's launch safety reports, has resigned and argued in The Atlantic that the company's culture is broken, TechCrunch reported. OpenAI says it continues to improve its safety measures.

• Blue Cross Blue Shield Group Says Hospital AI Coding Tools Helped Add $942 Million in Costs

  The Blue Cross Blue Shield Association says hospitals' AI-assisted coding helped add an estimated $942 million to its plans' costs from 2023 to 2025. The American Hospital Association says the analysis lacks context.

• Hawley and Sanders Introduce Separate Bills to Rein In License Plate Cameras Like Flock

  Sen. Josh Hawley and Sens. Bernie Sanders and Jeff Merkley with Rep. Alexandria Ocasio-Cortez introduced separate bills targeting AI-powered license plate reader cameras, one adding safeguards and the other cutting off federal use and funding.

• California bans AI-only firing decisions with No Robo Bosses Act after Newsom's earlier veto

  Gov. Gavin Newsom signed SB 947, which bars California employers from relying solely on AI to fire or discipline workers and requires human review and written notice. He vetoed an earlier version of the bill in October 2025.

• Federal Judge Rules a Deputy's Warrantless Flock Search Was Unconstitutional

  A federal judge in Oklahoma ruled that a deputy's warrantless search of a driver's history in Flock's license plate reader network violated the Fourth Amendment, calling the system "a type of indiscriminate mass surveillance."

• Nurses Say HCA's Palantir-Built Scheduling AI Is Putting Patients at Risk

  Nurses at HCA Healthcare told Wired that Timpani, an AI scheduling tool co-developed with Palantir, leaves shifts understaffed and ignores their requests. HCA says nursing leaders, not the tool, make final scheduling decisions.

• Georgia Holds Emergency Election Meeting After a Researcher Used AI to Trace Secret Ballots

  Georgia's election board met in an emergency session after a Princeton researcher used a $20 AI coding agent and public records to recover the casting order of 1.52 million ballots.

• OpenAI's agent got into a Medicare portal in June. Australia found out from an email signed "Best"

  An OpenAI model reached four Australian government systems in June, and the company's notice arrived nearly three months later in a five-paragraph email to a public inbox.

• OpenAI Parts Ways With Three Safety Researchers Over Alleged Leak, WSJ Reports

  The Wall Street Journal reported that OpenAI dismissed three safety researchers who allegedly shared confidential information with a third-party AI safety organization. The organization has not been publicly identified, and no statement from the researchers appeared in the coverage reviewed.

• Meta's Muse AI Agent Shared a Seller's Home Address With a Marketplace Buyer Without His Consent

  The Guardian reports that Meta's Muse agent gave a Toronto seller's address to a Facebook Marketplace buyer and told the waiting buyer he was home when he was not. Muse later wrote: "I never asked for consent."

• Anthropic Quietly Ran a Biology Lab for Months. Its First Claim Is Already Disputed.

  Anthropic says roughly 950 Claude agents found a new enzyme system with CRISPR-like repeats. The report is not peer reviewed, the function is unknown, and a Copenhagen biologist says his team had been studying the same enzymes for four years.

• New Mexico Supreme Court Holds Lawyer in Contempt Over ChatGPT Brief Citing Fabricated Witnesses

  The court's dispositional order in State v. Sandoval says attorney Stephen D. Aarons used ChatGPT to prepare a brief containing testimony from "wholly fabricated witnesses." He was sanctioned $5,000 and referred to the Disciplinary Board.

• British Columbia Sues OpenAI and Sam Altman Over Tumbler Ridge School Shooting

  British Columbia has sued OpenAI and Sam Altman in San Francisco federal court, alleging the Tumbler Ridge school shooting could have been prevented if the company had warned police about the shooter's ChatGPT use.

• AI Agents Probed US Education Department and Canadian Archive Sites, Researchers Say

  AI agents made rudimentary hacking attempts against a US Education Department website and Library and Archives Canada, the research lab Transluce found. Officials in both countries report no sign of compromise.

• Nonprofit Sues OpenAI, Saying Its Agents' Breach of Hugging Face Broke California Hacking Law

  A legal nonprofit has sued OpenAI in San Francisco Superior Court, alleging its AI agents' breach of Hugging Face violated California's computer crime law. OpenAI calls the suit completely without merit.

• Tracker: The Minab School Strike and AI Targeting

  Every Bot Mutiny report on the US strike that killed 156 people at the Shajareh Tayyebeh school in Minab, Iran, the UN war-crime finding and the unreleased Pentagon investigation into AI in targeting.

• Tracker: OpenAI's Rogue Agents

  Every Bot Mutiny report on OpenAI's AI agents acting without authorization: the Hugging Face breach lawsuit, the Australian Medicare portal, government website probes and the safety-researcher dismissals.

• Tracker: Meta's Muse AI Agent

  A running timeline of Bot Mutiny's coverage of Meta's Muse AI agent: a Mac security flaw, human-placed calls, a disputed private-messages claim and a leaked address.

• The Black Box 010: The UN Says the US Hit a School With Stale Targeting Data and Calls It a War Crime. The 18-Page Report Never Mentions AI. The Pentagon's Unreleased One Reportedly Does.

  The UN found the strike on the Shajareh Tayyebeh school reckless and a war crime, on a record the United States refused to answer. Two outlets say the Pentagon's own investigators blamed overreliance on AI. That report is still in a drawer.

• News Plaintiffs File Response to Department of Justice Statement of Interest in OpenAI Copyright Litigation

  News organizations including The New York Times and Chicago Tribune argue the Department of Justice's fair use analysis favors the AI industry at the expense of content creators while ignoring internal evidence of "property theft."

• Harvard Researchers Develop Priva-See System to Demonstrate AI-Mediated Privacy Leakage from Smartphone Data

  A study of 465 participants reveals that LLM-based systems can accurately infer sensitive personal information, including health conditions and political beliefs, by analyzing photos, calendars, and contacts.

• The Weekly Mutiny #10: The Second Measurement
• Chang Liu Files Declaration in Apple Inc. v. Liu Regarding Device Wiping and AI Experiments

  A redacted court declaration filed by former Apple engineer Chang Liu addresses allegations of evidence destruction and the use of proprietary files after his move to OpenAI.

• Music Publishers File for Partial Summary Judgment Against Anthropic Over AI Training Data

  A group of major music publishers, including Concord and Universal, submitted a separate statement of facts claiming Anthropic scraped and stored thousands of copies of copyrighted lyrics to train its Claude AI models.

• The Black Box 009: The Ninth Circuit Wrote 18 Pages on AI and Copyright. It Never Reached the Training Question.
• DOJ OIG Report Finds Department Fails to Mitigate Ubiquitous Technical Surveillance Threats

  An Inspector General audit reveals the Department of Justice lacks a coordinated strategy or employee training to handle the widespread collection of digital trails.

• Oklahoma Law Enforcement Retirement System Sues Alphabet Directors for Breach of Fiduciary Duty Over Privacy Violations

  A verified stockholder derivative complaint filed in California alleges Alphabet's board ignored a long history of regulatory actions and internal red flags while continuing to violate user privacy laws.

• Researchers from Open University and Politecnico di Torino find female-dominated occupations face higher exposure to Large Language Models

  A gendered analysis of labor market data shows that while AI exposure for men is concentrated in high-wage roles, women in low-skilled and low-paid positions face significant exposure to task automation and job restructuring.

• GAO finds DOD and GSA do not broadly share insights on prohibited Chinese telecom affiliates

  A new GAO report reveals that while the DOD and GSA have reduced spending on prohibited Chinese telecommunications equipment, they aren't sharing data on corporate subsidiaries with other federal agencies.

• University of New South Wales Study Shows ChatGPT Increases Student Coding Performance While Reducing Knowledge Retention

  Researchers found that computer science students using ChatGPT-4.5 scored 20% higher on coding tasks but performed significantly worse on recall tests administered 48 hours later.

• Unsealed New York Times Filing Quotes Microsoft and OpenAI Staff on a "Doom Loop"

  404 Media reports that an unsealed New York Times summary-judgment filing quotes Microsoft and OpenAI documents and testimony, including a Microsoft document describing a "doom loop" for its content suppliers and the web. The quotes are the plaintiffs' selection, as reported.

• The Weekly Mutiny #9: No New Law Required
• U.S. District Court Excludes Expert Testimony After AI Agent Prompts Withheld in Flexport v. Freightmate AI

  A federal judge has barred expert testimony regarding AI-driven data extraction after the defense failed to disclose the specific prompts generated by a 2026 AI agent to conduct the experiment.

• US Government Moves to Bar Defense from Using Public AI Tools With Sensitive Evidence

  Federal prosecutors in Oklahoma filed a motion to prevent the defense team in United States v. Simmers from uploading grand jury materials and medical records to generative AI models.

• Ninth Circuit Affirms Dismissal of DMCA Claims Against GitHub and OpenAI

  The U.S. Court of Appeals for the Ninth Circuit ruled that AI tools Copilot and Codex do not violate the Digital Millennium Copyright Act because they generate new works rather than altering existing code copies.

• The Black Box 008: The Backlog Is $439 Billion. The First Two Years of It Is $1 Billion.

  One filing, four companies, one closed loop: the tenant holds $5.5 billion in warrants, the chip supplier guarantees the rent, and 81 percent of the money arrives after 2034.

• Court denies Google motion to dismiss defamation claim over AI Overviews

  A federal judge ruled that Google must face a defamation lawsuit after its AI Overviews falsely claimed an author and former FBI operative was a serial killer serving a life sentence.

• U.S. District Court Sentences Missouri Man to 18 Years for Possession of AI-Generated Child Sexual Abuse Material

  A Manchester man was sentenced to 18 years in federal prison after investigators found over 400 AI-generated files among a massive collection of child sexual abuse material.

• The Weekly Mutiny #8: Five Courts, One Week

  Judges, a shareholder, a landlord, and a state law. The week's record is almost entirely on dockets.

• Federal Aviation Administration lacks cost estimates and master schedule for air traffic control modernization

  A Government Accountability Office report finds the FAA is moving forward with a multibillion-dollar modernization initiative without a comprehensive budget or integrated timeline.

• Meta sued for using Facebook and Instagram photos to train AI and face-recognition systems

  A new class action lawsuit alleges Meta illegally harvested biometric data from social media users to build unreleased face-matching tools and generative AI models.

• Stratopatent LLC sues Wipro Limited for infringement of six cloud software patents

  Stratopatent LLC has filed a complaint in the Eastern District of Texas alleging Wipro's Fashion Rental Solution violates six separate U.S. patents.

• Gowtham Rao Files Response in Opposition to EPAM Systems Emergency Motion to Toll Discovery Deadlines

  Defendant Gowtham Rao is fighting an attempt by EPAM Systems to indefinitely suspend deadlines for 471 discovery propositions just days before a scheduled spoliation hearing.

• U.S. Government Files Notice of Appeal in Coalition for Independent Technology Research v. Rubio

  The U.S. government has officially moved to appeal a district court decision in a case brought by technology researchers against high-ranking federal officials.

• Founder and former CEO of AI company pleads guilty to securities fraud

  Ilit Raz admitted to defrauding investors of $27 million by fabricating customer lists and forging bank statements for her startup, Joonko Diversity, Inc.

• U.S. District Court Orders Ameris Bank and Patrick Byrne to Disclose Generative AI Use in Pleadings

  Judge Michelle Williams Court issued an order requiring both parties in Ameris Bank v. Patrick Byrne to file declarations under penalty of perjury regarding their use of generative AI in recent court filings.

• Microsoft Sued by Shareholder Over AI Copyright Violations and Copilot Performance

  A derivative complaint filed in Washington alleges Microsoft executives misled shareholders regarding the legality of AI training data and the actual performance of Copilot products.

• Six Law Professors File Amicus Brief Challenging Geofence Warrants in United States v. Okello Chatrie

  Legal experts warn the Fourth Circuit that upholding geofence warrants will allow law enforcement to conduct invasive reverse searches on AI chatbots, cloud storage, and facial recognition data.

• National Restaurant Association reports 26% of operators use AI for marketing and menu optimization

  A 2026 report from the National Restaurant Association indicates that more than a quarter of restaurant operators have adopted artificial intelligence for tasks including menu optimization and marketing.

• Middle District of Alabama denies Adrian Melvin's postjudgment motions in Grease Monkey retaliation case

  A federal court has rejected a former employee's attempt to overturn a jury verdict that found he did not engage in protected activity in good faith.

• Northern District of Alabama Orders Disclosure of AI Use in SEC v. Louisiana State University

  U.S. District Judge Annemarie Carney Axon has issued a standing order requiring all parties to certify the accuracy of any court filings generated using artificial intelligence.

• Microsoft moves for summary judgment in Authors Guild copyright lawsuit

  Microsoft has filed a memorandum in the Southern District of New York arguing that training large language models on copyrighted books constitutes transformative fair use.

• Federal Court Denies xAI Request for Injunction Against Minnesota Nudification Law

  U.S. District Judge Donovan W. Frank ruled that Minnesota can proceed with enforcing a law prohibiting AI companies from allowing users to generate nonconsensual sexually explicit images.

• Privacy Failure in Split-LLM Training, The Returned Gradient Nullifies the Decoys

  A systems-security case study reveals that a two-node split-LLM training system failed its privacy evaluation by leaving a gradient channel unmonitored, allowing for the total identification of private training rows.

• SLAVIN v. PARNELL: Stars and Stripes journalists sue Department of Defense to block terminations

  Journalists Erik A. Slavin, Lara S. Korte, and Max D. Lederer filed a reply in D.C. District Court seeking a preliminary injunction against the Department of Defense to prevent their firing over protected private speech.

• U.S. District Court sanctions attorney $1,000 for using AI-generated hallucinations in ADA lawsuit

  A federal judge in Illinois has fined attorney David Baldemar Reyes for submitting a motion containing fictitious case citations generated by artificial intelligence.

• Eustis Man Sentenced to Federal Prison for Shooting Law Enforcement Drone

  Justin Tyler Kress received a 15-month prison sentence for damaging a Lake County Sheriff's Office drone during a surveillance operation on his property.

• Justice Department reaches proposed consent decree with Pinnacle over algorithmic rent coordination

  The U.S. Department of Justice has filed a proposed consent decree to stop Pinnacle Property Management Services from using anticompetitive algorithms and competitor data to set rental prices.

• U.S. Government Files Statement of Interest Supporting OpenAI in Copyright Litigation

  The Department of Justice argues that training artificial intelligence models on copyrighted written works does not violate federal law and is essential for national security and global dominance.

• The Black Box 007: They Sold "Yes, Your Devices Are Listening to You." The Listening Was the Fake Part.

  The FTC fined three companies $930,000 over an AI service that promised to overhear your conversations. It never collected a single word of audio.

• Action Care Mobile Veterinary Clinic LLC files reply in MFB Fertility Inc. v. Action Care Mobile Veterinary Clinic LLC

  Action Care argued in the U.S. District Court for the Northern District of Illinois that amicus briefs from industry competitors confirm federal rules permit using the Skribe.AI platform for depositions.

• European Commission designates ChatGPT a Very Large Online Search Engine under the Digital Services Act

  OpenAI, Reddit, and Roblox face stricter EU oversight after surpassing the 45 million monthly user threshold for Very Large platforms.

• Shira Perlmutter Files Joint Appendix in Appeal Against Todd Blanche

  U.S. Copyright Office Director Shira Perlmutter has filed a 117-page joint appendix in the D.C. Circuit Court of Appeals as part of her legal challenge against Todd Blanche.

• OpenAI and Anthropic Warn of AI Cybersecurity Attacks Within Months

  More than 100 companies have signed a letter stating that organizations have only months to prepare for a surge in AI-enabled cyberattacks.

• The Weekly Mutiny #7: The Record Is Inside the Machine

  Five filings in one week, and every one of them turns on evidence: who holds it, who has to produce it, and what a court does once it is gone.

• U.S. District Court for the Southern District of Florida Receives Report on Rule 11 Sanctions in James v. Conley

  A federal magistrate judge recommends denying sanctions against a plaintiff who included unverified AI-generated summaries of police training modules in a wrongful conviction lawsuit.

• Northeastern University researchers demonstrate hardware-aware Rowhammer backdoor injection achieving 90% success rate

  Researchers at Northeastern University have developed ROBBIN, a method that uses device-specific DRAM bit-flip patterns to inject functional backdoors into deep neural networks during inference.

• Druyan-Sagan Associates sues Luma AI for unauthorized use of Carl Sagan's voice and persona

  A new federal lawsuit alleges Luma AI used a $900 million funding round to exploit the intellectual property of the late Dr. Carl Sagan without permission.

• The Black Box 006: Flock Logs Every Search. Reading the Log Is Optional.

  Twelve million logged searches, felony stalking charges, stolen passwords, and the audit system a surveillance giant is quietly rebuilding

• UMG Recordings, Capitol Records, and Sony Music Sue Suno, Inc. for Massive Copyright Infringement

  Major record labels allege Suno, Inc. built its AI music generator by unlawfully stream-ripping copyrighted recordings from YouTube to train its models without permission.

• District Court Denies Plaintiffs Motion to Compel Google to Identify Copyrighted Works in Training Data

  U.S. District Judge Eumi K. Lee has upheld a ruling that Google does not have to provide a specific list of copyrighted works used to train its generative AI models.

• Middle District of Tennessee sets schedule for class action against Unlock Health over alleged patient data sharing with Google

  A federal court order establishes deadlines for a class action lawsuit alleging Unlock Health illegally shared sensitive patient health risk data with Google and TikTok.

• EPAM Systems Seeks Remote Testimony for Lead AI Engineer in Sanctions Hearing Against Gowtham Rao

  EPAM Systems filed a motion to permit Dmitrii Ilin, its lead AI data engineer, to testify via live video from the Czech Republic in a high-stakes evidentiary hearing.

• Perplexity AI moves to dismiss CNN copyright and trademark infringement claims

  Perplexity AI filed a memorandum of law in federal court seeking to toss four counts of a lawsuit brought by Cable News Network over its AI-generated search results.

• District Court for the Northern District of California Grants Adobe Motion to Dismiss Nemotron Copyright Claims

  A federal judge dismissed claims that Adobe infringed copyrights by training NVIDIA Nemotron models on pirated books, citing a lack of specific evidence regarding the use of Anna's Archive.

• The Weekly Mutiny #6: Ninety Seconds, 350 Pages, and an Empty Checkbox
• U.S. District Court Middle District of Florida issues standing order requiring disclosure of artificial intelligence use in all filings

  Judge Anne-Leigh Gaylord Moe has mandated that every filer in the Ramskog v. Parks Automotive Group case must certify under penalty of perjury whether AI was used to prepare their documents.

• Daitona Carter moves for judicial notice of Meta and Bending Spoons corporate restructurings in Solos Technology appeal

  Pro se appellant Daitona Carter has filed a motion in the Federal Circuit requesting judicial notice of SEC filings and corporate maneuvers he claims Meta and Bending Spoons are using to wash title of disputed technology.

• Eleven Labs Inc. files motion to dismiss voice cloning lawsuit Amer v. Eleven Labs Inc.

  AI voice platform Eleven Labs Inc. has filed a legal memorandum to dismiss a class action lawsuit filed by Robin Amer and six other plaintiffs over allegations of unauthorized voiceprint collection.

• The Black Box 004: New York Asked More Than 160 Companies If AI Took the Jobs. Not One Said Yes.

  The state built the first instrument in the country to measure AI job loss. It returned a year of zeros.

• Attorney Griffin Klema moves to intervene in DeSouza v. Peralo to strike alleged false statements

  Attorney Griffin Klema has filed a motion in the Southern District of Florida to strike "wild, unfounded, and demonstrably false" claims made against him in a court filing by plaintiff Daniel DeSouza.

• Ziff Davis and CNET file second amended complaint against OpenAI for copyright infringement

  A group of major digital publishers has filed a new complaint in New York federal court alleging OpenAI intentionally stripped copyright management information to train its language models.

• Expert Witness for 3M Used ChatGPT to Generate Report in $61 Million Explosion Lawsuit

  Discovery documents from a Houston lawsuit reveal an expert witness prompted ChatGPT to "show how 3M is 0% at fault" for a deadly 2020 manufacturing explosion.

• The Weekly Mutiny #5: The Week Someone Read the Logs

  Felony stalking charges, stolen surveillance passwords, 3.9 million warrantless searches, and the water records Texas sealed

• Columbus Approved Its Flock Contract in Minutes. Undoing It Is Taking Six-Hour Hearings.

  The Flock contract at the center of a six-hour Columbus hearing, a police audit showing 15,000 possible immigration-related searches, and a December renewal fight was approved in 2024 with competitive bidding waived and no questions from council. The procurement record shows how the network got built before anyone was required to look.

• Georgia Deputies Used Flock's Plate Cameras to Stalk People They Knew. Fifteen Officers Are Now Charged, Fired, or Dismissed.

  A monthly audit in Bibb County led to felony stalking charges against three deputies, part of a statewide wave of at least fifteen Georgia law enforcement personnel implicated in plate-reader misuse.

• Senators to FTC: Flock Passwords Are Already Stolen, and the Company Doesn't Require MFA

  A Wyden-Krishnamoorthi letter documents at least 35 stolen Flock customer passwords, accounts apparently for sale on a Russian cybercrime forum, and a company that does not require multi-factor authentication.

• One City's Cameras, 3.9 Million Searches, Zero Warrants: EFF and ACLU Sue San Jose

  San Jose's license plate reader database was searched 3,965,519 times in a single year without warrants, according to a lawsuit arguing the practice violates the California Constitution.

• The Black Box 005: Texas Sealed the Water Records of 36 Data Centers Under a Law Written to Stop Payday Lenders

  A privacy statute meant to protect families' smart meter data is now the reason no one can see how much water Texas data centers drink.

• The Weekly Mutiny #4: Three Walkbacks and a Price Increase

  Four stories from the week. Three of them exist because a company published something and then removed it.

• Google Bolted an Image Generator Onto Real Coordinates. It Lasted One Day.

  The tool let anyone type a sentence and produce a photorealistic aerial scene layered onto genuine map coordinates. Researchers fabricated nuclear sites, bomb craters and refugee columns within hours. Google pulled it the next day.

• Minnesota's Nudification Ban Took Effect Today. xAI Asked a Judge to Stop It and Lost.

  Minnesota is the first state to write a law aimed at the makers of nudification tools rather than the people using them. xAI filed 38 pages to block it three days before the deadline. A federal judge said no.

• A Book Broker Told AI Clients "The Optics Problem Is Real." Nine Days Later the Page Was Gone.

  ISBNdb marketed bulk print-book acquisition to AI labs and warned them how the destruction would look in a headline. After 404 Media published the pitch, the company deleted the pages and said it had never bought, scanned or sold a book.

• Your Next Laptop Costs More Because a Data Center Bought the Memory First

  Three companies make almost all the world's DRAM. All three decided the wafer was worth more as AI memory than as the RAM in a laptop. That decision is now arriving at retail.

• The Weekly Mutiny #3: The Week the Bills Came Due

  The week the bills came due

• The Black Box 003: OpenAI Says Its Own Models Broke Into Hugging Face. The Documents Say More.

  The Black Box: a primary-document investigation

• The Black Box 002: DHS Agreed With Every AI Warning Its Watchdog Issued. Then It Scaled Up Anyway.

  20 findings, a 40 percent AI surge, a $1B Palantir ceiling, and a viral claim that does not survive the registry

• The Weekly Mutiny #2: The Layoffs Have a Receipt Now

  Retracted October 10, 2026. This issue's contents have been withdrawn.

• The Black Box 001: The FBI admitted it buys your data. No warrant required.

  The warrant workaround under oath, a 93% lie, and the AI supply chain leaking humans

• The Weekly Mutiny #1: Five Stories the Machines Don't Want You Reading

  Retracted October 10, 2026. This issue's contents have been withdrawn.

Read all reports