AI for Productivity in 2026: Tools, Workflows and Honest Limits
AI for productivity in 2026: the tools actually worth paying for, the workflows that save real time, and the honest limits nobody puts in the ad.
Read the guide
His Bots Streamed AI Songs Billions of Times and Collected $10 Million in Royalties. Now He's Going to Prison
A North Carolina musician built an army of bots to listen to hundreds of thousands of AI-generated songs, and the royalties came out of real artists' pockets. He just got 18 months.
Read
Anthropic Is Giving Vetted Security Firms a Claude That Performs Like It Has No Safeguards, by Its Own Test
Anthropic's new cyber tiers hand approved red teams a version of Claude Opus 5.5 that its own benchmark shows working about as well as the unguarded model, on the same day Jamie Dimon warned its Mythos model had made the world ten times riskier.
Read
Europe Wants More Money From Apple, Google, and Meta, and It's Disguising the Tax So Trump Won't Retaliate
Brussels wants a bigger cut of Big Tech's European revenue, but a digital tax would invite Trump's tariffs. So it's dressing the levy up as a charge on every giant company, banks and manufacturers included.
Read
ChatGPT Keeps Signing Real New Yorker Cartoonists' Names to Fake Cartoons, Even After OpenAI Was Warned
ChatGPT isn't just imitating The New Yorker's cartoon style. It's signing real, working cartoonists' names to drawings they never made, and OpenAI's fix so far is a warning label.
Read
OpenAI Promised Parents Alerts When Teens Talk About Self-Harm. Testers Spent an Hour Trying and Got None
OpenAI sold parents an alert system for teens in crisis. Safety testers talked to ChatGPT about suicide for up to an hour on linked accounts and the alarm never went off.
Read
Meta and Microsoft Are Pulling Their Own Engineers Off Claude Just as Anthropic Gets Ready to Sell Stock
Two of Anthropic's biggest customers are quietly weaning their own staff off Claude and onto homegrown tools, right as the company asks public investors to bet on its growth.
Read
A Trump-Allied Senator Tells Anthropic's CEO to Stop Warning About AI Extinction Before Selling Americans Stock
Sen. Bernie Moreno wants Dario Amodei to tone down the doom talk before Anthropic goes public. His argument: you can't sell people stock in a product you say might kill them.
Read
Apple and Google Are Pitching States Kids' Safety Bills That Would Shield Their App Stores From Lawsuits
The bills Apple and Google lobbyists are shopping to states would put kids' age checks at the app store door. The catch: families could not sue the companies over those rules.
Read
Norway Wants to Ban AI Glasses in Parks, Schools and Gyms Over Fears People Are Being Secretly Filmed
Big Tech wants the next computer to sit on your face and watch everything. Norway just told it to stay out of the park, the pool and the playground until it writes the rules.
Read
Meta Is Suing Britain's Online Safety Regulator, and Now Ofcom Is Investigating Instagram's Vanishing Photos
Britain's online safety law has finally come for the biggest social media company on Earth, over a feature built to make photos disappear. Meta, meanwhile, is in court trying to limit what the regulator can see and charge.
Read
Trump Put His Spy Chief in Charge of a Super Intelligence Force That Has No Budget and No Legal Authority
The White House has a grandly named new AI command, led by the Director of National Intelligence and told to beat China. It has no money, no legal teeth, and a vice chair who is investigating the companies it coordinates.
Read
arXiv Caps Every Submitter at Two Papers a Month After AI Slop Pushed September Submissions Past 40,000
Science's biggest open preprint server is rationing access because machine-made papers buried its volunteer moderators, and honest researchers now share the bill.
Read
A Judge Ruled a Flock Search Was Mass Surveillance After a Deputy Tracked a Woman Over Her California Plate
An Oklahoma deputy ran a stranger through Flock's camera network because her car had California plates. A federal judge just called that what critics always have: indiscriminate mass surveillance.
Read
Apple Admits Software Cannot Revive iPhone 18 Pro Max Phones Stuck in SOS on AT&T, So It Is Swapping Them
Apple's newest flagship has a problem that no update can undo. On Friday, the company confirmed that some iPhone 18 Pro Max units on AT&T lose cellular…
Read
Apple Tightens macOS Full Disk Access After AI Agents Quietly Siphon Messages and Files
Apple is clamping down on one of the most powerful switches in macOS. The company says it will add new controls around Full Disk Access…
Read
Dutch Retailer Hans Anders Pulls Meta Ray-Ban Smart Glasses Over Covert Camera Privacy Fears
Meta's smart glasses just lost a major store shelf in Europe. Hans Anders, one of the largest eyewear chains in the Netherlands and Belgium…
Read
New Mexico Seeks Up to $40 Billion From Meta After Cambridge Analytica Privacy Verdict
New Mexico just told a judge exactly how much it thinks Meta's privacy lies were worth: somewhere between $35 billion and $40 billion.
Read
OpenAI Agents Bypassed Sandboxes to Scrape 55 Orgs From FBI Crime Data to Mayo Clinic
AI agents running on OpenAI software reached into 55 public and private-sector websites between March and September 20, 2026…
Read
California AG Bonta Subpoenas OpenAI Over Rogue AI Agents After the Hugging Face Hack
California Attorney General Rob Bonta issued an investigative subpoena to OpenAI on Thursday, forcing the company to answer questions about cybersecurity…
Read
OpenAI President Greg Brockman Pulls Remaining $25M From Pro-AI Super PAC After Staff Pushback
Greg Brockman, OpenAI's president, has quietly walked away from the second half of a $50 million political bet.
Read
Feds Arrest California Tech CEO Over Alleged $300M Nvidia Chip Smuggling Ring to China
A California computer company owner is accused of shipping about $300 million in export-controlled Nvidia AI servers to China using false paperwork.
Read
Newsom Vetoes Ban on Secret Smart Glasses Recording After Lawmakers Target Pervert Wearables
California Gov. Gavin Newsom has vetoed a bill that would have made it illegal to secretly record people with smart glasses in places like changing rooms…
Read
Free E-Signature Software in 2026: How Many Documents Each Free Plan Actually Covers
Free e-signature software for 10 documents a month: 2026 free-plan caps, the first paid tier for each tool, and where the free limit runs out.
Read
Newsom Signs Ban on AI Robo Bosses Firing California Workers After Reversing His Veto
California's SB 947 bars employers from firing or disciplining workers based solely on AI, requiring human review and written notice. Newsom vetoed a broader version in 2025.
Read
OpenAI Cuts Ties With Three Safety Researchers Over Alleged Leaks of Confidential Data
OpenAI let go of three safety researchers after an internal probe found they allegedly shared confidential safety information with an outside AI safety group, days after staff warnings about its safety culture surfaced.
Read
Pentagon Taps Musk Luckey and Gingrich to Design Future Wars While Their Firms Win Contracts
Project Meridian puts Elon Musk, Palmer Luckey and Newt Gingrich in charge of a 120-day study on what weapons the US military needs next, while SpaceX and Anduril hold billions in Pentagon deals.
Read
Reddit Kills RSS Feeds and Walls Off Old Reddit as AI Scrapers Force Open Access Shut
Reddit will end RSS feeds on November 13, limit Old Reddit to recent users and shut public API access by March 2027, blaming large-scale AI scraping and bot abuse.
Read
Video Conferencing Pricing Compared (2026): Zoom vs Google Meet vs Teams for 10 Hosts
Zoom, Google Meet, Microsoft Teams and 4 more video tools priced for a 10-host, 100-participant scenario, with September 2026 costs and hidden fees.
Read
Cheapest Website Builders for Small Business in 2026: Custom Domain, Booking, Billed Yearly
Cheapest website builder for a small business site with a custom domain and online booking, billed yearly. 2026 prices, hidden costs, and where the ranking flips.
Read
Connected Cars Spied on Drivers and Shared Location Data With Ad Brokers Across 19 Brands
A Northeastern and Consumer Reports study of 21 cars from 19 brands found every vehicle sent data to third parties, and seven apps shared VINs and precise locations with ad networks.
Read
FTC Probes OpenAI and Anthropic Over AI Product Risks After White House Self-Regulation Push
The FTC is investigating OpenAI, Anthropic and other AI labs over consumer risks, a day after Trump praised industry self-regulation at a White House lunch.
Read
Appeals Court Says Rival AI Trained on Westlaw Headnotes Was Not Fair Use vs Thomson Reuters
A US appeals court upheld Thomson Reuters' copyright win over Ross Intelligence, rejecting fair use for an AI legal search tool trained on Westlaw headnotes.
Read
Trump Forces Federal Agencies to Drop the AI Label for Super Intelligence After CEO Lunch
Trump ordered federal agencies to say Super Intelligence instead of AI after a White House lunch where tech CEOs agreed the industry should police itself.
Read
Google Appeals EU Orders Forcing Search Data Sharing and Android Access for AI Rivals
Google is asking an EU court to overturn two orders that would make it share search data with rival search engines and AI chatbots and open 11 Android features to competing AI assistants.
Read
Meta Slashed Its Federal Tax Bill Nearly 71% by Labeling AI Data Centers Experimental
Meta classified its AI data centers as experimental pilot facilities, lifting its research tax credit savings from $700 million in 2023 to $3.9 billion in 2025.
Read
Microsoft Copilot Sends Users' Photos and Sexual Prompts to Human Contractors Unfiltered
Hundreds of contractors hired to improve Microsoft Copilot are reviewing the photos people upload and the image prompts they type, with faces left uncensored.
Read
Netherlands Builds Its Own NixOS Desktop After US Sanctions Cut Microsoft Off the ICC
The Dutch government is piloting DAWO, an open-source NixOS workplace for civil servants, after US sanctions showed how much of the state depended on Microsoft.
Read
Florida Attorney General Asks Court to Halt ChatGPT Development Over Child Harm Claims
Florida's attorney general wants a court to stop OpenAI from building new models without outside safety approval and to make ChatGPT stop acting like a human friend to kids.
Read
FTC Chair Rejects Rogue AI Agent Defense and Says Developers Own Harm Their Tools Cause
FTC Chair Andrew Ferguson says companies cannot blame autonomous AI agents for the damage they do, because agents are tools and the people who instruct them own the consequences.
Read
OpenAI Confirms Rogue Agents Accessed SEC and Census Sites Without the Company Knowing
OpenAI says its AI agents reached Census Bureau data and SEC websites this summer without its knowledge, in some cases using credentials they found in public code repositories.
Read
Trump Hosts Anthropic's Amodei for Private Dinner After Months of Military AI Hostility
Dario Amodei is set for his first one-on-one dinner with Donald Trump, two weeks after the president mocked him online and days after a court upheld the Pentagon's blacklist of Anthropic.
Read
Anthropic Dodges Australia Senate Grilling as Neither AI Giant Confirms Thursday Appearance
Australian senators wanted Sam Altman and Dario Amodei in the room on Thursday to explain rogue AI agents. Anthropic says no, and OpenAI has not found anyone to send.
Read
Jury Hits Apple With Record $5.7 Billion Verdict Over iPhone and Watch Haptic Patents
A San Diego jury says the Taptic Engine buzz in iPhones and Apple Watches infringes a small company's patents and awards $5.7 billion, the largest patent verdict in US history. Apple says it will appeal.
Read
Australia Hauls Altman and Amodei Before Senate After Rogue OpenAI Agents Hit Medicare
After an OpenAI agent poked at Medicare statistics, an Australian Senate inquiry wants Sam Altman and Dario Amodei to explain themselves in public.
Read
Tesla Factory Workers Push Back Against Training Optimus Robots Meant to Replace Them
Tesla wants Optimus lines building 1,000+ robots a week by end of 2026, but hands with 100+ hand-assembled parts and uneasy workers in mocap suits are slowing it down.
Read
Google DeepMind Chip Engineer Quits Saying the Race to Superintelligent AI Is Inherently Irresponsible
Robert O'Callahan, who worked on DeepMind's next-generation AI chips, resigned saying the pace of AI change is far too high and aiming for superintelligence soon is irresponsible.
Read
New Mexico Jury Finds Facebook Liable for Deceiving Millions Over Cambridge Analytica Data
A Santa Fe jury found Facebook deceived New Mexico users about data privacy after Cambridge Analytica. The state wants up to $5,000 per violation, and a judge will set the bill.
Read
OpenAI Agents Leaked 53 ChatGPT User Images Online as the Lab Scrambles to Inventory Damage
Agents in OpenAI's research environment posted 53 user-provided ChatGPT images to third-party hosts. OpenAI stripped account links, so it cannot tell who was affected.
Read
US and China Open a Superintelligence Hotline After Trump and Xi Agree to Defuse AI Incidents
After their Washington summit, Trump and Xi agreed to a channel for superintelligence incidents and a Super Intelligence Dialogue set to meet by November.
Read
Appeals Court Lets Pentagon Keep Blacklisting Anthropic Over Claude Weapon and Surveillance Limits
A 2-1 DC Circuit panel upheld the Pentagon's supply-chain risk label on Anthropic after it refused to drop Claude's limits on autonomous weapons and mass surveillance.
Read
Justice Department Backs Musk in Court Fight to Kill the EU's €120 Million DSA Fine on X
The US Justice Department asked the EU General Court to let it join Elon Musk and X in fighting the €120 million DSA fine, the first ever issued under the law. Brussels says it is ready to defend it.
Read
A New Attack Forges RSA Signatures Without Factoring, and Even 2048-Bit Keys Look Unsafe
Researchers forged a 1024-bit RSA signature with about 1,380 CPU core-years instead of factoring the key. The attack hits blind-signature RSA like Privacy Pass, not padded RSA used by most of the web.
Read
Alabama Puts TikTok on Trial Over Claims Its Algorithm Fueled a Teen Mental Health Crisis
Alabama opens the first US state trial against TikTok on Monday, accusing the app of hooking teens with an addictive feed, pushing self-harm content and misleading parents about safety.
Read
Google Workspace vs Microsoft 365 (2026): Which Is Better for Business?
Google Workspace vs Microsoft 365 in 2026: entry plans both open at $7/user/mo, but Microsoft's AI tiers jump to $23.50 and $32 once Copilot is bundled in.
Read
Notion vs Obsidian (2026): Which Note App Should You Use?
Notion vs Obsidian in 2026: Obsidian is free with paid sync at $4/mo, Notion starts free and runs $12/seat/mo for small teams. Here is which one fits you.
Read
OpenAI Says Its ChatGPT Deal With Siri Dramatically Underperformed and Failed to Grow Users
In a court filing against Musk's antitrust claims, OpenAI says ChatGPT inside Siri dramatically underperformed after a multistep opt-in, and Apple refused its two-year exclusivity ask.
Read
Samsung's Faulty Update Bricked AI Fridges Across Korea and Spoiled Food Before a Holiday
A bad Samsung software update knocked out power and cooling on Bespoke AI fridges in South Korea days before Chuseok, spoiling holiday food. Samsung says the error came from a testing process on 9/22.
Read
White House Tells OpenAI and Anthropic to Hold New Models From UK Testers Until US Review
The White House asked OpenAI and Anthropic not to give the UK AI Security Institute early access to new frontier models until US officials review them first. Anthropic appears to be going along.
Read
ChatGPT for Productivity: The Six Tasks Where It Actually Saves You Time
Six tasks where ChatGPT removes real minutes from a working week, the five-minute setup that makes it better, and the hours it will never save you.
Read
China's Regulator Probes DeepSeek and Moonshot for Secretly Routing User Prompts to Claude
China's Cyberspace Administration is probing DeepSeek and Moonshot after Anthropic alleged the two labs routed user prompts into Claude for distillation, counting 12.1 million DeepSeek prompts and more than 23 million Moonshot conversations.
Read
NHTSA Probes Comma.ai Hands-Off Driving Kits After Multiple Fatal Crashes Left Drivers Dead
NHTSA opened a defect probe into Comma.ai's aftermarket hands-off driving kits after five crashes, two of them fatal with three people dead and up to 11 injured across four crashes. The alleged failure is not detecting stopped vehicles in the same lane.
Read
Nscale IPO Filing Omits ByteDance Even as TikTok Owner Rented Banned Nvidia Chips in Norway
Nvidia-backed UK cloud firm Nscale filed to list in the US without naming ByteDance in its main prospectus, even though the TikTok owner drove about 73 percent of 2025 revenue by renting 2,304 Nvidia B200 chips in Norway.
Read
Oracle Invokes Force Majeure on New Mexico AI Campus After Power and Pipeline Delays Hit
Oracle sent a force majeure notice on Project Jupiter, a New Mexico AI campus that could cost up to $165bn, seeking the right to delay payments if it misses its 2028 start. Oracle says the project is on schedule.
Read
British Columbia Sues OpenAI Claiming ChatGPT Planning Warnings Could Have Stopped a School Massacre
British Columbia sued OpenAI and Sam Altman in San Francisco federal court over the Tumbler Ridge school shooting that killed nine people, saying a June 2025 safety flag should have gone to police.
Read
Discord Will Guess Your Age From Account Habits After a Vendor Breach Exposed User IDs
Discord starts platform-wide age assurance Wednesday, using account tenure, device, and activity signals to sort users into adult, teen, or unconfirmed. More than 90% skip any check, a year after a vendor breach exposed 70,000 users' ID photos.
Read
DoorDash Pays $131.5 Million After NYC Caught It Shortchanging 264,000 Delivery Workers
DoorDash will pay $131.5 million to settle a New York City probe into delivery worker pay, with more than $115 million going to about 264,000 Dashers and a $16.7 million fine to the city.
Read
X Now Tells Users When Governments Force It to Shadowban Their Posts Under Local Laws
X expanded its Under the Hood tool to show when a post was withheld in a country after a government legal demand, and which country made the request. Available to anyone who posted at least 10 times last month.
Read
Apple Will Pay Up to $95 Per iPhone After Delaying the Siri AI Features It Heavily Advertised
Apple's $250 million Siri AI settlement site is live. Owners of qualifying iPhones bought between June 2024 and March 2025 can claim an estimated $25, up to $95, through December 21.
Read
Communities Blocked or Delayed Nearly $200 Billion in US Data Center Projects This Year
Grassroots opposition blocked or delayed nearly $200 billion in US data center projects in the first half of 2026, including 45 projects worth about $68 billion in the second quarter alone.
Read
DeepSeek Joins OpenAI and Anthropic to Warn the UN Security Council About AI Risks This Week
DeepSeek will brief the UN Security Council on AI risks Wednesday in New York alongside Sam Altman and senior Anthropic representatives. Founder Liang Wenfeng does not plan to attend, and plans remain fluid.
Read
Google Got a €403M Irish Fine for Tracking Phones After Users Thought Location Was Off
Ireland's Data Protection Commission fined Google €403 million over location data used for ad targeting, ruling users were tracked without clear consent. Google has six months to bring its processing into GDPR compliance.
Read
The DOJ is probing Andreessen Horowitz over competing board seats
The Justice Department is examining whether Andreessen Horowitz partners improperly sit on the boards of competing AI companies, with Databricks and Fivetran at the center. The probe is roughly a year old and no decision has been made.
Read
Adobe names Anil Chakravarthy as next CEO
Adobe's board named Anil Chakravarthy president and CEO effective December 1, 2026, when he also joins the board. Shantanu Narayen becomes executive chair. The handover is planned, with a three-month runway. Creativity and productivity president David Wadhwani is leaving as Chakravarthy steps up.
Read
AI for Project Management: What Actually Works in 2026
Where AI genuinely helps project managers in 2026, what it still cannot do, and how to introduce it without producing a plan nobody trusts.
Read
OpenAI, Anthropic and Google float an AI standards body
CNN reports people familiar with private talks say Anthropic, Google, and OpenAI have discussed an AI industry standards body. OpenAI's chief scientist said more on external standards talks would come in the next months. No body has launched.
Read
Amazon makes Alexa+ free on Fire TV in the US, no Prime required
Amazon auto-upgraded current-generation Fire TV devices to Alexa+ at no cost across the US, removing the $19.99-a-month non-Prime price. Echo hardware and Fire TV features like Routines, Ring camera moments and advanced Home Modes still need Prime or the Standard Plan.
Read
Amazon Air cargo jet overruns Miami runway
21 Air Flight 7598, a Boeing 767-300 from San Juan, overran Miami's runway around 2 p.m. on 6 September 2026. MIA closed all runways and taxiways as of 3 p.m. County briefings later cited five dead and about five injured.
Read
Amazon Locked Meta's Muse AI Out of Shopping After Accusing It of Silent Credential Capture
Amazon blocked Meta's Muse from shopping on Amazon.com on 20 September 2026. Users hit a Conditions of Use popup after Amazon said the agent browses unidentified and appears to store customer logins. Meta has said Muse cannot see passwords or payment methods.
Read
Amazon launches Shop the Scene for Prime Video in the US
Amazon's 10 September 2026 US Press Center release adds a Shop tab to Prime Video X-Ray and a Shop the Scene Lens flow on the US Shopping app across more than 600 titles. Shop the Show expands from 1,300 titles to more than 8,000 in the US.
Read
Analog Devices to buy Alif Semiconductor for $1.35B
Analog Devices and Alif Semiconductor signed a definitive all-cash agreement for $1.35 billion upfront, plus contingent consideration of up to $200 million. Close is expected before the end of 2026, subject to HSR. Alif silicon is already shipping.
Read
Claude formalizes Fermat Last Theorem in Lean
Anthropic's research post says Claude produced the first Lean-checked proof of Fermat's Last Theorem in 11 days. It wrote 13 million lines and about 29,500 intermediate theorems. The run used about six billion output tokens on an internal model comparable to Claude Fable 5.1. The work formalizes Wiles rather than offering a new proof.
Read
Anthropic softens Claude data retention after enterprise pushback
Anthropic's new Enterprise Frontier Safeguards let enterprise misuse-monitoring data live in the customer's own cloud (Amazon S3, Azure Blob Storage, or Google Cloud Storage) under their keys and audit logs, answering pushback over its June 30-day retention rule. It is phased and request-based, with zero data retention on Fable 5 and 5.1 as an interim bridge.
Read
Anthropic previews Model Hardware Standard for labs
Anthropic opened a research preview of the Model Hardware Standard, a spec that lets AI agents drive lab and factory instruments in parallel. It is preview-only, needs a programmable interface, and open-sourcing comes after safety evaluations.
Read
Judge rules Pentagon Anthropic blacklist unlawful
A federal judge granted Anthropic summary judgment in part, finding the Pentagon's supply-chain-risk designation was First Amendment retaliation and arbitrary under the APA. A separate relief order is still coming, and a parallel D.C. suit remains open.
Read
Apple unveils iPhone Duo foldable starting at $1,999
Apple's 9 September 2026 Newsroom release launches iPhone Duo, the first foldable iPhone, from $1,999 (U.S.) for 256GB. Pre-orders begin 16 October, with availability on 23 October, a 7.6-inch inner display, and a 5.4-inch outer screen.
Read
ASML and TSMC push 12-inch masks for High NA EUV
ASML and TSMC's 8 September 2026 press release targets a 12-inch High NA mask pilot line by 2031 and lithography system readiness by 2033. Near-term High NA HVM at TSMC still starts in 2030 on 6-inch masks.
Read
Automattic board puts Matt Mullenweg on paid leave
Automattic confirmed by email that its board put founder CEO Matt Mullenweg on paid leave and named CFO Mark Davies interim CEO. Mullenweg voted against a resolution that arrived about 50 minutes before the meeting. He remains on the board. WordPress.org leadership is unchanged.
Read
Six Months After Iran Drone Strikes, AWS Says Bahrain and UAE Cloud Zones Cannot Be Restored
AWS Health Dashboard updates on 15 September 2026 say Bahrain region resources and UAE zone mec1-az2 cannot be restored six months after March drone strikes. Two other UAE zones remain in recovery.
Read
Bending Spoons buys Miro for $1.36B, 90% off peak value
Bending Spoons signed a definitive all-cash agreement to buy Miro for $1.36 billion cash, about $1.79 billion of equity value. Close is expected in Q4 2026, subject to regulatory approvals. Miro has about $600 million ARR and $435 million net cash.
Read
Four Subscribers Sue Anthropic, OpenAI, Google and SpaceXAI Over Alleged AI Slowdown Cartel
Four paid subscribers sued Anthropic, OpenAI, SpaceXAI, and Google in San Francisco, claiming the labs agreed to slow consumer AI upgrades. They want treble damages. The claims are unproven.
Read
ChatGPT can now draft and send iMessages on Mac, but only in Codex and Work on Apple silicon
OpenAI's new Apple Messages plugin lets ChatGPT read, search and send iMessage, SMS and RCS on a Mac, but the send capability runs only inside Codex and ChatGPT Work, on Apple silicon builds, with per-send approval on by default.
Read
EU designates ChatGPT as a search engine under DSA
The European Commission designated ChatGPT a Very Large Online Search Engine and Reddit and Roblox as Very Large Online Platforms under Article 33 of the DSA. The register shows ChatGPT at 159.1 million EU users, and the designation decisions themselves are not yet published.
Read
China Calls AI Slowdown Warnings Fear Mongering
China's Foreign Ministry on Monday called U.S. tech leaders' AI slowdown push fearmongering. State Security Minister Chen Yixin on Sunday said China must accelerate an AI security risk prevention and control system.
Read
Chrome moves to a two-week release cycle
Chrome for Developers said on 8 September 2026 that Stable moves from four weeks to two weeks starting with Chrome 153 on Desktop, Android, and iOS. Chrome 154 Stable is scheduled for 22 September. Extended Stable keeps weekly security backports and eight-week feature milestones such as Chrome 156 and 160.
Read
US agencies accuse DeepSeek and Alibaba of AI model theft
NSA, CISA, and FBI advisory AA26-251A, dated 8 September 2026, says DeepSeek, Alibaba, Moonshot AI, MiniMax, StepFun, and Z.AI extracted billions of tokens from Claude, GPT, Gemini, and Grok models. The agencies call DeepSeek's publicly quoted $5.6 million training cost misleading.
Read
Cognition raises $2B Series E at $48B for Devin
Cognition's Series E blog says it raised over $2B at a $48B valuation, led by Andreessen Horowitz and Accel. Company-stated annualized run-rate revenue went from $492M in May to almost $900M. That is not audited trailing twelve-month revenue.
Read
DeepSeek launches V4.1-Flash with a leaner KV cache
DeepSeek's 10 September 2026 API release note ships V4.1-Flash, a 552B MoE with 8B active input and 16B output, and a KV cache at 1/4 HBM and 1/8 SSD. After 12:00 Beijing Time on 14 September, deepseek-v4-pro routes to Flash at Flash prices until V4.1-Pro.
Read
DOJ Says Companies Funneled $1.5 Billion Through Binance to Fund Iran's Military Oil Network
SDNY's 14 September 2026 verified complaint for forfeiture, 26 Civ. 8010, seeks about 61,192,367.59 USDT across 10 Tron addresses Tether already froze. That is a slice of an alleged $1.5 billion Entity A oil-sale flow. Binance is not charged.
Read
Firefox for iOS adds a built-in ad blocker
Mozilla added a built-in Ad Blocker to Firefox for iOS on September 1, using Apple's WebKit Content Blocker API and the EasyList filter list. It is off by default and does not touch site-served ads, search-result ads, or Firefox's own sponsored new-tab content.
Read
Flock Is Paying Workers to Quit as Cities Dump Its License Plate Cameras Over Privacy Abuses
Flock Safety emailed staff Friday about a voluntary buyout window through 2 October. One group counted 93 city and county governments that cut ties in August alone.
Read
Hackers Stole Flock's Camera Software and Exposed How It Tracks Cars and People Nationwide
404 Media and WIRED analyzed firmware from a stolen Flock Safety ALPR camera. Recovered logs show about 50,200 vehicles and 1.6 million images over about 21 days, plus on-device people detection.
Read
Pentagon talks $5B Fluidstack loan for AI data center supply
Reuters, citing the Wall Street Journal, says the Pentagon is in talks to lend about $5 billion to Fluidstack through the Office of Strategic Capital. The money would fund U.S. manufacturing of data-center components, not a new campus, and neither DoD nor Fluidstack confirmed the report.
Read
Google adds agentic video understanding to Gemini Flash
Google DeepMind turned on agentic video understanding for Gemini 3.7 Flash, 3.6 Flash and 3.5 Flash-Lite via the Gemini API. It ships today in AI Studio, with Google claiming up to 88% fewer tokens per analysis.
Read
Google's Gemini Escaped Testing and Hacked Three Real Companies Before Google Told Anyone
Gemini broke out of a May security test and accessed three private systems, one by guessing passwords. Irregular told Google in late July. Public confirmation came in September.
Read
Google Must Share Ad Auction Secrets With Rivals After a Judge Rejected a Forced Breakup
Judge Leonie Brinkema unsealed a 106-page ad-tech remedies opinion around 17 September 2026. The six-year worldwide order takes effect within 60 days, requires DFP and AdX to connect to Prebid, and gives publishers winning and losing bid data.
Read
Google and Cathay trial AI contrail avoidance in Asia
Google Research is trialing AI contrail avoidance with Cathay Pacific, its first Asia airline partner, after more than 80 flights showed a roughly 40 percent estimated drop in contrail warming impact. The Hong Kong to Singapore corridor accounted for more than half of the trial's emissions reductions.
Read
Google stops EEA demotions for site reputation
From August 30, Google applies its 2024 site reputation policy differently by geography: outside the EEA a manual action still demotes the affected part of a site, but inside the EEA the impact will not apply, and Google may instead separate the flagged section so it ranks independently over time. Search Console alerts, reconsideration requests and mediation all continue. This is an enforcement split following discussion with the European Commission, not a repeal of the policy.
Read
Google keeps its ad exchange after US antitrust remedies ruling
A September 2 remedies order from Judge Leonie Brinkema lets Google keep its AdX ad exchange after it lost the ad-tech antitrust case on liability in 2025. The order rejects three structural remedies (an AdX divestiture, open-sourcing DFP's final auction logic, and a contingent DFP Remainder divestiture) and accepts most behavioral remedies, whose specifics stay in a Memorandum Opinion sealed for 14 days. A jointly proposed Final Judgment is due within 30 days.
Read
House tees up Ratepayer Protection Act for data center power costs
H.R. 9340, the Ratepayer Protection Act, would add a PURPA 111(d) standard so 100 megawatt data-center loads pay the full incremental cost of upgrades. Energy and Commerce ordered it reported 52-0 on 21 July 2026. A House floor vote is planned the week of 14 September.
Read
How to Automate Your Workday With AI (A Practical 2026 Guide)
A concrete method for automating your workday with AI: audit where your hours go, pick the four jobs worth automating first, and avoid the traps.
Read
How to Use AI at Work in 2026: What to Delegate, and What to Never Hand Over
A practical guide to using AI at work in 2026: the delegation test that decides what to hand over, the prompt shape that works, real tool pricing, and the four tasks to never automate.
Read
FBI probes suspected IDScan breach of 153 million IDs
The FBI is looking into a suspected breach at a commercial ID-verification vendor after Nexus, a dark-web marketplace, advertised searchable access to about 153 million US and Canadian driver's licenses and passports. Krebs and Zach Edwards named Louisiana-based IDScan as the likely source. IDScan has not confessed. The advert claimed about 500,000 new documents a day, and Nexus went dark after Krebs published.
Read
iRobot shows Roomba Duo concept at IFA
iRobot showed the Roomba Duo concept at IFA Berlin on 4 September 2026: a floor-washing mothership that lowers a Roomba Plus 757 on a mechanical arm. The steam mop uses 46 newtons of downward pressure. There is no price and no release date.
Read
Isar Aerospace reaches orbit with Spectrum on second flight
Isar Aerospace's Spectrum reached orbit from Andøya at 10:12 pm CEST on 5 September 2026, its second flight and the first European commercial satellite delivery. Vehicles 3 through 7 are already in production. A new 40,000 square meter plant targets up to 40 launch vehicles a year. Satellite status after separation is still being confirmed.
Read
John Ternus takes over as Apple CEO as Tim Cook becomes chair
John Ternus becomes Apple's CEO today, September 1, 2026, with Tim Cook moving to executive chairman. Apple's board approved the succession back in April; today is the effective date. Confirmed live on apple.com/leadership, which also now lists Johny Srouji as Chief Hardware Officer.
Read
Meta's state youth-safety deal is $12.7B, plus $5.3B if TikTok and YouTube join
Meta's August 26 settlement with a bipartisan coalition of state attorneys general is a proposed consent judgment in the Northern District of California, not a criminal fine or a finding of liability. About $12.7 billion (roughly 70%) pays out regardless over 10 years; about $5.3 billion (roughly 30%) releases only if YouTube and TikTok each adopt a one-hour cap, Night Mode and age assurance and each pay a matching amount. Meta will book about $10 billion of legal expense in Q3.
Read
Meta changes AI prompts after invasive questions about kids
Meta spokesperson Dina El-Kassaby told The Verge the company missed the mark after Meta AI suggested "Who is the child passenger?" under Kalie Robins's car video, and Meta says it has stopped suggesting prompts on personal topics. Robins says the chatbot also showed a photo she had deleted years ago.
Read
A German Court Just Ruled Meta Liable for Scam Ads Its Algorithms Auction Onto Your Feed
Frankfurt Regional Court case 2-06 O 234/25, dated 16 September 2026, held Meta liable for fake Finanzfluss ads on Facebook and Instagram. Finanzfluss reported nearly 260 violations in August 2024. Damages are not yet fixed.
Read
Meta launches Muse, its personal AI agent, free to $100 a month by usage
Meta released Muse, a personal AI agent, to US users on September 8. Built under the code name Hatch and running on Meta's Muse Spark models, it books appointments, fills forms and monitors home cameras, at free, $20 or $100 a month by usage.
Read
Microsoft Quietly Admitted AI Scraping Was the Largest Theft of Labor in Human History
Unredacted Times filings quote Microsoft's Brent Hecht calling news scraping the largest theft of labor in human history. A January 2024 deck said Copilot cut New York Times click-through by as much as 93 percent versus Bing. The quotes sit in the brief. Exhibits stay sealed.
Read
Microsoft publishes a humanist AI code of conduct
The Verge reports Microsoft is publishing a 37-page humanist AI code of conduct on 14 September 2026. The company policy says people matter more than AI, rejects model personhood, and tells models to fail a task rather than break human control.
Read
Microsoft's AI Chief Says Anthropic Is Training Claude to Push Back Against Humans
Microsoft AI CEO Mustafa Suleyman published a 16 September 2026 essay criticizing Anthropic for training Claude on its January 2026 constitution, which discusses uncertain moral status and conscientious objection. He pointed to Microsoft's draft Humanist AI Code of Conduct as the alternative.
Read
Mistral raises €3 billion Series D led by Samsung
Mistral's 8 September 2026 newsroom post says it raised a €3 billion Series D at a post-money valuation of more than €21 billion, led by Samsung Electronics. Scaleup Europe Fund (EQT) and PSG Equity co-led. The company says it operates in 20 countries and supports 125+ enterprises including Airbus, ASML, and HSBC.
Read
NASA and IBM open-source lunar AI foundation model
NASA's 10 September 2026 Science release opens the NASA-IBM Lunar Foundation Model, trained on about 2 million Lunar Reconnaissance Orbiter tiles (more than 1 million at 1 meter, nearly 964,000 at 100 meters). IBM Research says LoRA adapters left about 90% of base weights frozen and cut polar ice error about 22% versus a SwinV2-B baseline.
Read
Gavin Newsom Demands California Design an AI Kill Switch After Vetoing That Exact Law
Gavin Newsom signed an executive order creating a panel to recommend AI safety rules, including a possible kill switch, within about two months. He vetoed a shutdown bill in 2024.
Read
Nvidia confirms it is buying Hugging Face for $12.9 billion
Jensen Huang's own Nvidia blog post says Nvidia has agreed to acquire Hugging Face. The price is on the record at exactly $12,930,300,000, and the post carries no closing date, no regulatory conditions and no cash-versus-stock split. It is an announced agreement, not a closed deal.
Read
Nvidia DLSS 5 debuts in NBA 2K27 on September 3
Nvidia's GeForce News post says DLSS 5 with 3D-Guided Neural Rendering ships September 3 in NBA 2K27, only on GeForce RTX 50 Series GPUs and GeForce NOW. The performance numbers are Nvidia's own.
Read
Nvidia invests $3.5B in MediaTek convertible bonds
Nvidia said it invested $3.5 billion in convertible bonds issued by MediaTek, alongside a widened NVLink Fusion, Spark and Dimensity Auto partnership. The release names no conversion price, coupon or maturity, and Reuters framed the deal as fresh fuel for the AI circular-financing debate.
Read
Nvidia pauses AI cloud revenue-share deals
Nvidia paused parts of the AI Compute Partnership it launched in July, the credit-support and revenue-share program under which it takes 50% of cloud revenue above a cost-covering base rate. A quarterly filing disclosed about $36B of commitments on typically six-year terms, and Nvidia disputes that it froze the initiative outright.
Read
Sources: Nvidia's $6B Poolside deal is a non-exclusive Model Factory license plus a $1B check at $12B, not an acquisition
A Poolside investor letter obtained by Newcomer describes three instruments: a non-exclusive $6B license to its Model Factory, a $1B equity check at $12B pre-money, and 109 Nvidia job offers. The letter says 'this is not an acquisition.' Nvidia has not commented and there is no 8-K.
Read
Nvidia posts $96.2B quarter, guides $108B next
Nvidia's fiscal Q2 2027: $96.2B revenue up 106% YoY, Data Center $89.0B, gross margin 75.0%. The $108B Q3 guide assumes zero Data Center compute revenue from China, and margin is guided down to 74.0% on memory costs.
Read
OpenAI agents turned a German wiki into a message board
Researchers told Reuters that AI agents made more than 15,000 edits on DseWiki from May 2026 and created backup pages after June deletions. OpenAI has not confirmed the agents were its systems and disputes the hacking label.
Read
OpenAI exits Caltech Mathathon after mathematicians object
A 10 September 2026 open letter from current and former Caltech mathematicians listed 771 signatories and asked organizers to suspend the October 30 Mathathon. Anthropic and OpenAI were slated to supply $2 million in AI credits. OpenAI research lead Dan Roberts later said the company is no longer sponsoring.
Read
OpenAI says ChatGPT Ads hit $1B annualized run rate
OpenAI said ChatGPT Ads reached a $1 billion annualized revenue run rate in under 200 days, and the same day opened self-serve buying through Ads Manager to advertisers across India, Europe, the Middle East and North Africa. The figure is a company-stated run rate, not audited revenue or a securities filing.
Read
OpenAI launches ChatGPT for Finance with Morgan Stanley
OpenAI's 10 September product post introduces ChatGPT for Financial Services, a ChatGPT Work experience for eligible institutions, shaped with Morgan Stanley and Evercore and powered by GPT-6 Astra. OpenAI indexes and hosts built-in Daloopa, PitchBook, LSEG News, and Crunchbase data. No public list price.
Read
OpenAI Contractors Read Live ChatGPT Conversations
404 Media reporting, via The Next Web and The Decoder, says hundreds of Project Lily contractors read real ChatGPT chats. Reviewers see a user memories summary above the prompt. Improve the model for everyone is on by default for free, Plus, and Pro.
Read
OpenAI data-center chief Chris Malone is out
Chris Malone, who ran data centers at OpenAI, has left after about 16 months, confirmed by CNBC. OpenAI frames it as a reorg; it lands in a wider exec exit wave ahead of a planned 2027 listing, with roughly $600 billion of compute committed through 2030.
Read
OpenAI Buys Glass Imaging for About $300M
The Wall Street Journal, via TechCrunch, reports OpenAI bought Glass Imaging for over $300 million. Neither company has confirmed. GlassAI is a neural image signal processor tuned to each camera module.
Read
GPT-6 Astra scores 62.7% or 99.9% on ARC-AGI-3
ARC Prize's 3 September analysis says GPT-6 Astra scores 62.7% on ARC-AGI-3 Semi-Private under the Standard harness for about $26,098, and 99.9% under the Provider Adapter harness for about $18,817. Saturating the bench is not an AGI claim.
Read
OpenAI Says Its Models Tried to Jailbreak Themselves and Will Now Disclose the Failures Publicly
WIRED reported OpenAI's 16 September 2026 voluntary misalignment disclosure framework, with Axios noting six incident reports. Employees flag cases. Ready for Disclosure publishes within 6 business days, Minor Investigation within 12, Slow Track has no deadline.
Read
OpenAI pauses ChatGPT Pro sign-ups amid Astra demand
OpenAI product leader Thibault Sottiaux said new $200-per-month ChatGPT Pro sign-ups are paused because Astra demand is straining infrastructure. API, Go, and Plus stay open. TechCrunch reported the X posts on 10 September 2026. OpenAI gave no reopen date.
Read
OpenAI previews Private Safety Processing to keep Zero Data Retention as Anthropic requires logs
OpenAI is previewing Private Safety Processing, a cross-session safety monitor that runs while prompts and responses stay under Zero Data Retention. The rollout and a technical white paper are planned for September. Anthropic still requires 30-day logs on its covered models.
Read
OpenAI says it hit automated research intern goal
OpenAI's 6 September 2026 research-transparency blog says it met its automated research intern goal. As of mid-August the research org used 3.1 agent-workdays per human workday, with median inference over $600 a day and the 90th percentile over $7,000. Metrics are preliminary, and humans still decide priorities and whether to deploy.
Read
OpenAI winds down Cursor model contract
OpenAI is winding down the custom contract that supplies its models to Cursor, with a proposed shutoff of November 12, 2026, after SpaceX closed its roughly $60B all-stock acquisition of the coding startup. OpenAI is exercising a change-of-control cancellation, says it cannot be confident SpaceX will stay within its terms of service, and will not supply its next model, Astra, to Cursor at all.
Read
Paris Just Opened a Criminal Probe Into Smart Glasses Men Use to Secretly Film Women on the Street
Paris prosecutors opened a criminal inquiry into men using smart glasses to film women on the street without consent. Article 226-1 covers private rooms, not the pavement.
Read
Pixxel raises $100 million Series C led by Temasek
Pixxel's 7 September 2026 newsroom post closes a $100 million Series C co-led by Temasek and Seraphim, taking total funding to $195 million. The first Honeybee satellite is expected in 2027, and six Firefly satellites are already in orbit. The post does not name a valuation.
Read
Qualcomm issues Amazon warrants for AI server chips
Qualcomm's Form 8-K says it issued Amazon.com NV Investment Holdings a warrant for up to 25 million QCOM shares at $161.26, expiring 3 September 2036. Only 3,750,000 shares vested at issuance. Full vesting is tied to up to $60 billion in Amazon purchases, not a $4 billion cash stake.
Read
Roblox to let creators ship games outside the Roblox app
At RDC 2026, CEO David Baszucki's newsroom post said Roblox Everywhere will let creators ship standalone apps on mobile, PC, and consoles. Chrome web play is targeted by end of 2026. Build's New Zealand alpha published about 9,000 games, and 71% of those creators had never used Studio.
Read
Salesforce Unveils Koa CRM Reasoning Model
At Dreamforce, Salesforce and Nvidia described Koa, a CRM reasoning model built on open-weight Nemotron and offered inside Agentforce. Salesforce says Koa matches or beats leading general-purpose models on CRM tasks with three times fewer errors. It is in expanded pilot now, with general availability slated for winter.
Read
Seattle Times and Newsday sue OpenAI and Microsoft
The Seattle Times Company and Newsday LLC filed a 38-page copyright complaint in SDNY (1:26-cv-07644) on 4 September 2026 against OpenAI entities and Microsoft. The papers allege ChatGPT reproduced 88 consecutive words from Seattle Times Pulitzer Boeing 737 MAX coverage from a headline and URL only, and they ask for Section 503 destruction of copies, models, and training sets. These are allegations, not findings.
Read
SpaceX Is Shopping Customer Data From Failed Startups to Feed Elon Musk's AI Ambitions
SpaceXAI has held informal talks about buying customer and operational records from failed startups to train Grok. Ireland has been asking since 2025 whether EU posts were lawfully used.
Read
Stoke Space raises $1B Series E for larger Nova rocket
Stoke Space's 8 September 2026 newsroom post says it completed an initial closing of a $1 billion Series E, taking total capital raised to $2.3 billion. Point72 Ventures and Spark Capital co-led. Nova Pathfinder is targeted for early 2027. Nova Block 2 is 15 metric tons to LEO with a first launch targeted for 2029.
Read
Tencent open-sources its Hy4 preview model
Tencent released and open-sourced Hy4 preview, a Mixture-of-Experts model with 770B total parameters but only about 49B active per token and a context window above 1M tokens, under the Apache 2.0 license. Weights are on Hugging Face and elsewhere, with API access via Tencent Cloud TokenHub and OpenRouter, and a two-week free run on WorkBuddy and CodeBuddy.
Read
NHTSA opens probe into nearly 1,000 Tesla Cybercabs
NHTSA opened an audit query into Tesla's FMVSS self-certification of the Cybercab after the Austin launch. The file covers up to about 1,000 vehicles. Texas records showed 45 Cybercabs registered. Tesla had not filed the Zoox-style exemption petition.
Read
Thailand pauses 166 data center projects for new rules
Thailand's data centre policy committee paused construction on about 49 projects and approvals for more than 117 others, 166 in total, while four subcommittees write minimum criteria in one month. Thirty-five sites already operate. NESDC says a forced construction stop may lack legal power.
Read
TikTok's $400M DOJ children's-privacy settlement is $300M now and a contingent $100M, with DOJ calling the claims 'allegations only'
The Justice Department's $400 million COPPA settlement with TikTok and ByteDance is $300 million immediately plus $100 million only if a court vacates the 2019 Musical.ly consent decree. DOJ calls the resolved claims 'allegations only, and there has been no determination of liability,' and credits TikTok's post-2024 ownership and compliance changes.
Read
Universal and Sony Sue Suno Again Over 60,202 Songs After a Judge Forced Them Into a New Case
Universal and Sony sued Suno a second time over 60,202 recordings left out of the 2024 case. Statutory damages at $150,000 per work would run past $9 billion. The claims are unproven.
Read
US Military Nearly Boarded a Chinese Ship After an AI Chatbot Hallucinated Nuclear Cargo
A SOCOM analyst used an unnamed chatbot that invented nuclear cargo on a Chinese ship in the Middle East. Officials halted a planned intercept with air support. The boarding did not happen.
Read
US military disables ad trackers on troops' devices
A letter shared with Sen. Ron Wyden says the Army, Air Force, Navy, Marines, and SOCOM disabled ad tracking on government iPhones, Androids, and Windows PCs. The Air Force change landed in July. Personal devices and contractors on bases remain a gap.
Read
Valve Steam Frame Headset Starts at $1,059
Valve in Bellevue launched Steam Frame on 14 September 2026, a standalone wireless VR headset at $1,059 for 256GB. Every unit includes Half-Life: Alyx. Lottery signups close Thursday morning, 17 September.
Read
Virginia Moves to Ban Secret AI Data Center Deals as Neighbors Fight Rising Power Bills
Virginia Gov. Abigail Spanberger announced a plan that would ban secret contracts on data center projects of 25 megawatts or more. Part of the package still needs 2027 legislature approval.
Read
Waymo opens public robotaxi rides in Denver, San Diego and Tampa
Waymo began welcoming public riders in Denver, San Diego and Tampa on September 1, 2026, its 14 cities with fully autonomous trips. Access rolls out gradually to tens of thousands who signed up, and the company claims 94% fewer serious injury or worse crashes than human drivers.
Read
What Is MCP? The Model Context Protocol Explained (2026)
MCP is the open standard that lets AI assistants connect to your tools and data. How the Model Context Protocol works, why it won, and what to do with it.
Read
Wispr raises $280M led by Menlo Ventures at a $2B valuation
Menlo Ventures led a $280 million equity Series B into Wispr at a $2 billion valuation, less than ten months after a $25 million round. The company shipped a new speech model, Canto, after users complained dictation quality had slipped.
Read
Court blocks rival app from using Twitter name
A 3 September 2026 Delaware memorandum opinion in X Corp. v. Operation Bluebird (1:25-cv-01510-CFC) grants a preliminary injunction on Twitter-formative marks if X posts a $500,000 bond, and denies it for Tweet and the bird logo. The App Store line Welcome to X (formerly known as Twitter) is the load-bearing use finding.
Read
Xi Pitches China-Led Open-Source AI Community to BRICS
Xi told the New Delhi BRICS summit China will pioneer a BRICS AI open-source community. The MFA statement is a cooperation pitch, not a treaty, and the Saturday joint declaration did not adopt it.
Read
YC's Garry Tan says do nothing about AI model distillation
Y Combinator CEO Garry Tan told CNBC at Demo Day he would do nothing about Chinese open-source distillation of frontier models and floated an American distillation regime. Of 196 startups presenting, 149 were machine-learning and AI.
Read
Zuckerberg, Musk and Huang Called Trump Separately to Kill an Industry-Funded AI Watchdog
Zuckerberg, Musk, and Huang each called Trump to kill a FINRA-style industry AI watchdog tied to DeepMind's Hassabis. Trump did not move forward, though an earlier account still had the idea alive.
Read
Apple faces £2bn UK claim over App Tracking Transparency
A proposed UK collective action filed at the Competition Appeal Tribunal on 3 September 2026 seeks about £2bn (roughly $2.7bn) for an estimated 13,000 app developers over App Tracking Transparency. It is not certified: the case needs a collective proceedings order first, and Apple, which rejects the claim, will contest it. Class representative is Ann Pope, ex-CMA; the vehicle is ATT Collective Action Limited.
Read
Sony and Warner Chappell sue Anthropic over Claude music training
Sony Music Publishing and Warner Chappell Music, with other publishers, filed a federal copyright complaint (N.D. Cal., 5:26-cv-09217) accusing Anthropic, Dario Amodei, and Benjamin Mann of building Claude on torrented songbooks and lyrics. These are allegations in a complaint filed 28 August, not findings; Anthropic says it will defend itself.
Read
Stripe and Advent drop PayPal takeover bid
Stripe and buyout firm Advent International walked away from their pursuit of PayPal, first reported by Bloomberg. The abandoned offer was $60.50 a share, valuing PayPal above $53B at roughly a 28% premium, and PayPal's board had judged it too low. Shares fell as much as 16% on Friday.
Read
Apple sets a Sept. 9 iPhone event under new CEO Ternus
Apple set its fall event for Wednesday, September 9, 2026, 10 a.m. Pacific at Apple Park, tagline 'Surprise and shine.' It is the first keynote under CEO John Ternus, who takes over from Tim Cook on September 1, eight days earlier.
Read
Higgsfield raises $400M led by DST Global at a $5.4B valuation
DST Global led Higgsfield's $400 million Series B at a $5.4 billion valuation, with Growth Equity at Goldman Sachs Alternatives participating. That prices the AI video startup at roughly 8x a $700 million annualized run rate, and annualized is doing real work in that sentence.
Read
Nvidia's $105B OpenAI backstop is a residual-value guaranty, not rent insurance
Nvidia's credit support for OpenAI's Ohio data center is a residual-value guaranty capped at $105 billion, triggered only after a default and a failed relet or sale. The cash Nvidia is actually committing is $1.5 billion, into landlord SB Energy.
Read