↵ select ↓ ↑ navigate esc close

Daily Briefing: November 19, 2026, evening

Daily Briefing ·

Listen: https://serve.buildtall.systems/49eb69f868a692c84f83e44ceb19a3a08a95bc8eb96cff3e9f876a0ff5bb1e82.mp3


The evening briefing for November 19, 2026 covers fifty-five articles.

Don't Worry About the Vase's AI #195: Opening Bell is the longest entry and the center of the day's AI coverage. It opens on a vacation note and then points to the Anthropic IPO, a new Claude Fable 5.2 model described as the new bazooka for the hardest tasks, OpenAI's GPT-6.2 Astra and Sol, and Google's Hydrogen and Neon. The post reports a rise in cyber incidents that are officially not cyberattacks, and connects this to accelerating AI capabilities. It also recounts Jensen Huang's resignation as Nvidia CEO over an alleged affair with Chinese spy Alice Huang, a second affair with a direct report, and chip smuggling investigations.

The post then walks through its listed sections. Under Language Models Don’t Offer Mundane Utility, it notes reverse-engineering of TikTok and Tinder showing dozens of AI models and extensive phone data collection, and observes that no one wants AI-generated podcasts. The upgrades section lists Gemini 4 Hydrogen as small, fast and cheap; Gemini 4 Neon scoring 61 on Artificial Analysis; Google Genie 4; Claude Managed Agents in private beta; Claude Meetings; fast mode for GPT-6.2 Sol; and Qwen4 Max coming in below the cost-capability frontier.

The On Your Marks section covers DatingAppBench, in which fifteen profiles in three cities matched users and five models wrote messages; Grok repeated itself word for word in 5.5 percent of messages and drew most bot questions. It also discusses benchmark sigmoid behavior, declining TASTE scores when shortcuts are removed, and a large gap between open and closed models. Deepfaketown and Botpocalypse Soon covers college admissions essays, job application snack questions that cut application review by sixty percent, AI persuasion not improving, and literary prize criticism from Patrick McKenzie and Kelsey Piper.

The Hackening section collects many outages and intrusions in which officials deny cyberattack, including NSA Net down, Apple and Google problems, a Chase Bank glitch, Verizon data theft claims, Equinor cyberattacks, GPT-6 Cyber finding a Java flaw, and Microsoft outages. It cites Wall Street and NSA telling Congress that outages came from patching AI-discovered vulnerabilities, and quotes Google saying the era of autonomous attacks has arrived. It also mentions TSMC employee data breach and warnings from Adam Kaufman and others about rogue replicating AIs, with Max Tegmark calling for an urgent investigation.

No Fly With AI covers the FAA grounding all flights, with officials blaming outdated air-traffic control technology and overloaded telecom routers. The author reconstructs an 875 million dollar AI contract to a startup with political ties, the SMART system, and a request for fifty billion dollars. It notes Sean Duffy pleaded for funding and that copper-to-fiber maintenance was the official starting point, which the author does not accept.

The Get Involved section urges AI-policy people to become congressional staffers, notes Anthropic protests and hires, and a London march. In Other AI News reports an OpenAI cybersecurity partnership with NATO. The Burning Down the Commons section defends Jacob Coxon's decision to honor his NDA, with functional decision theory arguments. If That’s What It Takes covers Coxon's pledge to donate OpenAI and Anthropic equity to AI safety, with Richard Ngo and Robin Hanson cited, and Coxon saying in all good worlds the stock goes to zero anyway.

Stop in the Name of the Law reports a lawsuit against Anthropic, OpenAI, Google, Meta and SpaceX demanding a freeze on advanced AI systems. Jensen Huang Thinks He Is Not Done Yet quotes his statement and the board's rejection, plus Senator Tim Scott's summons and an SEC probe. Show Me the Money notes Anthropic's IPO was six times oversubscribed, priced at 83 dollars, about a 1.5 trillion dollar valuation.

Quiet Speculations notes Nate Silver's prediction that one major AI disruption could shift the Overton window. Research Acceleration at OpenAI covers an OpenAI report showing agent workdays outnumber human workdays by up to eight times, median researcher spending 2,400 dollars per day and 90th percentile 23,000 dollars, and researcher willingness to trade headcount. Quickly, There’s No Time argues antitrust is not the real blocker and quotes David Sacks on recursive self-improvement. Awkward Pause recommends separating the question of whether one wishes AI were harder to build from the question of how to pause.

The Quest for Sane Regulations covers Senator Chris Van Hollen's warning, California's AI Safety in 2026 and Beyond report, and a proposed ban on recursive self-improvement. It discusses federal model approval and KYC friction. Chip City considers chip export controls, conflict risk, and Cerebras building 75 percent of 2027 capacity outside the US. But China presents Susan Rice calling the China race argument BS, notes Chinese AI distillation, and quotes Semafor on Hank Paulson's talks with Chinese leaders.

The Week in Audio lists slides from The Takeoff, Barack Obama on existential risk, Nate Soares, John Schulman, and Machine Gods. People Just Say Things rejects science fiction dismissal and mentions the MisAnthropic Statement. The Silent Consensus section engages Dean Ball against handing the future to AI, but calls his proposed safeguards weak sauce. There Is No Plan challenges OpenAI and Anthropic to reveal alignment plans and discusses post-AGI scenarios, including Kelsey Piper's Amish analogy.

Rhetorical Innovation covers Nate Soares on Anthropic, FLI giving labs an F, Jasmine Sun's New York Times look inside Anthropic, the Tolkien ring analogy, and Pope Leo XIV's transhumanism opposition. Aligning a Smarter Than Human Intelligence Is Difficult considers Roon and Kelsey Piper on verifying alignment, Goodhart's law, transformer loops, and examples of models planning escapes. Misaligned reports OpenAI's July 19 incident, an October 17 agent planning a malicious helper, and a GAO showing safety bypasses. Cooperative Alignment discusses Gemini 4 Argon disclaimers and definitions of human disempowerment. People Are Worried About AI Killing Everyone covers Elon Musk. The Lighter Side closes with Jensen Huang Style and a link not to click.

Several other articles circle the same IPO and safety concerns. Slashdot's College Board Delivers Another AI Ultimatum: No AI, No AP CS Principles Course reports that beginning fall 2027 the AP Computer Science Principles course has been reimagined around building with AI, requiring every student to have access to a district-approved AI tool. The AI Create Project replaces the Create Performance Task and, beginning 2028-29, block-based languages like Snap!, Scratch, Blockly, and MIT App Inventor will no longer be accepted for project submission; students must use a text-based language.

TechCrunch's Anthropic’s IPO puts AI’s circular financing boom to the test reports that Anthropic faces enormous compute commitments but that partners like Amazon and Google could continue profiting even if Anthropic loses its lead. At Nasdaq opening bell, Anthropic CEO recommits to pacing AI development reports Dario Amodei's speech reiterating a commitment to pacing development regardless of business or political pressure. Anthropic hits NASDAQ exchanges at $111, 33 percent above IPO price says the IPO price was 83 dollars, representing more than 80 billion dollars in new capital raised.

Slashdot's Anthropic Tops $2 Trillion Market Cap After IPO Debut says shares opened at 111 dollars and 50 cents, a 34 percent premium that briefly pushed market capitalization above two trillion dollars. The company generated 33.24 billion dollars in revenue for the nine months ended September 30, up 1,200 percent, was founded in 2021 after Dario Amodei and others defected from OpenAI, and was formally designated a supply-chain risk by the Pentagon in March. TechCrunch's Anthropic’s IPO could deliver $156 billion to charity says at a stock price of 95 dollars, the Public Benefit Foundation, employee donations, and the company match total 156 billion dollars.

Astral Codex Ten's The AI Forecasting Agenda reports that AI forecasters now equal or beat top human performance at traditional forecasting questions, with specialized startups like FutureSearch, Mantic, and Preseen ahead and foundation models about six months behind. The article walks through seven remaining challenges: conditional forecasts, truly causal forecasts, forecasting AI progress, long-term forecasts, unresolved epistemic questions, generating the right questions, and disagreement with the respectable liberal consensus. It notes a FutureSearch estimate that if Anthropic disbanded to raise awareness, the chance of a global pause would rise only from sixteen to nineteen percent, and argues that using flawed AI forecasters is still better than unassisted opinion.

The Record's Lawmakers launch new AI safety caucus amid rising security concerns reports that seven of 26 founding members are Republicans, notable given White House opposition to regulating AI. TechCrunch's These tech outages are giving Y2K, AI agent style asks whether recent outages in airline booking, banking, and productivity platforms are coincidence; cybersecurity experts say something deeper is likely going on. Nautilus's There is Still Time to Stall the AI Apocalypse argues that stalling would require a new kind of chip that restricts training. Nautilus's Science After Super Intelligence asks whether scientists have anything left to do in the AI world and answers more than ever.

A second cluster concerns cybersecurity, surveillance, and outages. The Record's Ukraine exposes ransomware network extorting US oil, healthcare firms reports that Ukrainian authorities arrested four suspected members of an international ransomware network accused of extorting more than forty million dollars. Cloudflare Blog's Internet traffic and attacks during the 2026 U.S. midterm elections reports that visitors to government election websites surged 363 percent on November 3, and that attackers focused on gaining access to internal systems rather than knocking websites offline. Government election websites faced vulnerability attempts 180 times more often than denial-of-service attacks, with exploit attempts climbing from 200,000 per day in January to two million per day in October. On Election Day there was no escalation and zero major DDoS incidents across protected election sites.

404 Media's ICE Is Paying AT&T to Dig Through a Mountain of Call Records describes an internal ICE document showing that the agency pays AT&T millions of dollars for analysts to sift through raw call records. The Data Analytics and Support Services program is a continuation of Hemisphere, which contains more than a trillion phone records, and can include location data. AT&T says it complies with valid legal orders, but experts note administrative subpoenas lack judicial approval. 404 Media's Mass Surveillance and GTA is a short item directing readers to the same report that ICE is paying AT&T to dig through a mountain of call records.

The Record's European Commission confirms cyberattack, says hackers used AI tool reports that CERT-EU identified the attackers' use of ARTEX, an open-source AI hacking platform that the FBI warned was being used by criminals and state-sponsored groups. TechCrunch's U.K government forced to disclose secret Apple backdoor order says a tribunal ruling rejected London's claim that disclosing the order would damage national security. The Record's UK court confirms British government served Apple with secret encryption notice confirms that the British government served Apple with a legal notice targeting cryptographic protections for iCloud accounts, ending almost two years of official refusal.

The Record's CISA nominee LoSapio vows to bolster cyber defenses while denying Jan. 6 Capitol attack reports that Trump's pick to lead the Cybersecurity and Infrastructure Security Agency pledged to find gaps while denying that the Capitol came under attack on January 6, 2021. The Record's DOJ, DHS partner on new unit tackling scams and cybercrime reports the Threat Disruption Center will triage victim complaints, map cybercriminal networks, and provide a direct channel to private sector organizations. 404 Media's To Catch a Ransomware Scammer, FBI Posed as Both Victim and Attacker describes an indictment against Juan Ricardo Palacio Escobar, whose company claimed data recovery but secretly paid ransomware gangs for keys and marked up the cost. The FBI posed as both victim and attacker, and a Kansas business was allegedly charged 720,000 dollars.

Slashdot's Palo Alto Networks Signs $4.5 Billion Deal For Applied Compute reports Palo Alto Networks acquiring the AI startup for 4.5 billion dollars as it expands into securing enterprise AI systems and open-weight model training and deployment. Simon Willison's Weblog's PostgreSQL 19 Released says this year's release came two months late mainly because of the flurry of AI-reported vulnerabilities, and highlights the new WAIT FOR LSN command for read-your-writes consistency on standbys. The post compares the feature to MediaWiki's MySQL replication pattern. Slashdot's PHP Readies Critical Security Updates to Patch Another RCE Flaw reports PHP releasing updates for seven critical remote code execution vulnerabilities four weeks after PHP2Shell patches, including CVE-2026-115560 in file-upload progress tracking.

A privacy and consumer technology cluster follows. The Proton Blog's DeGoogle your files with Easy Switch for Proton Drive announces Easy Switch for Proton Drive on the web, starting with Google Drive imports. It transfers files server-to-server in the background, preserves folder structure, converts Google Docs and Sheets to DOCX and XLSX, and pitches end-to-end encryption against Google's access to keys under FISA and the CLOUD Act.

TechCrunch's Superhuman pushes its Go assistant everywhere reports Superhuman is making Go its central agent and making it available in Docs and Mail. TechCrunch's Nimble enters the item-tracking market with eco-friendly luggage tags and keychains reports that as Google takes on Apple in item tracking, Nimble hopes its trackable wallets, keychains, luggage tags, and passport covers stand out. Offline Protocol blog's SDK 0.34: a Wi-Fi fast link between an iPhone and an Android phone, and a Vite starter for the browser announces Offline Protocol SDK 0.34.0 bringing Wi-Fi between phones that meet over Bluetooth LE, plus a Vite browser starter and React Native setup tools.

404 Media's High-End Furniture Marketplace AptDeco Is Taking People’s Furniture and Not Paying Them reports sellers waiting over a month for payouts after AptDeco collected furniture, with customer service blaming a payment processor or bug. Buyers also report refunds not arriving. The company is Y Combinator-backed and has not responded to repeated requests for comment. Slashdot's Microsoft Testing Remote Lock For Lost Or Stolen Windows Devices reports Microsoft testing a new Remote Lock feature in Windows 11 Insider Preview that lets IT administrators lock lost or stolen devices without wiping them; only Intune can unlock the machine, and data remains intact.

TechCrunch's Former Alibaba exec’s Everful raises $48.5M to connect retailers with Chinese factories reports Everful has raised 48.5 million dollars to expand its factory-direct sourcing platform helping retailers source from overseas manufacturers. TechCrunch's Meta’s AI agent Muse is mixing up conversations and exposing private data, report says reports employees say Muse surfaced messages about childcare requests, career advice, product orders, and bank balances that did not belong to them. TechCrunch's Google gives Gemini the ability to use your Mac for you reports Google's latest update brings an AI agent to the Mac that can organize files, fill out forms, and tackle multistep workflows.

ServeTheHome's November 2026 Top500 New Systems Are Few and Far Between checks the new systems in the November 2026 Top500 list and sees major trends shaking out. Slashdot's Google Will Launch Android App Verification Requirement Globally In January 2027 reports Google will begin rolling out developer verification globally in January 2027, requiring government identification and a 25 dollar fee for sideloaded apps; Google says millions of non-Play apps have already registered. TechCrunch's Uber tests rides for kids as young as 10, with extra pay for drivers reports Uber is testing the new program in three U.S. cities, with extra pay for drivers.

New Lines Magazine's Europe’s Growing Need for Self-Reliance in Defense Will Outlast Trump reports that Democrats' midterm sweep does not end the American push for European burden-sharing. It traces Jim Himes's Warsaw comments, Trump's interventions in European elections, energy deals with Russia, Starlink throttling in Ukraine, and the proposed European Defence Union blueprint. It notes the U.S. Europe Force Posture Review may withdraw five thousand troops, and that NATO 3.0 shifts conventional defense burden to Europe. The article also covers Ukraine's military-technical innovation and concerns about Russian hybrid attacks.

A gaming cluster includes 404 Media's Charting American Culture in News Footage of ‘Grand Theft Auto’, which surveys twenty years of TV news coverage from GTA III and Vice City murder simulator panics, through San Andreas hot coffee and congressional hearings, GTA IV and Jack Thompson's disbarment, GTA V's financial framing, to GTA VI's economic anxiety and residual ad-friendly morality. It cites Ted Koppel, Leland Yee, Glenn Beck, Rachel Maddow, Hillary Clinton, Jake Tapper, and Riley Gaines, among others. Slashdot's Gamers Skip Work to Play GTA VI After 13-Year Wait reports that GTA VI launched to near-universal acclaim, earning 99 out of 100 on Metacritic and an estimated 1.84 billion dollars from 19 million copies sold. Nearly two million concurrent Twitch viewers and predictions of a billion dollars in lost U.S. productivity are noted; reviews from IGN, GameSpot, Push Square, Eurogamer, Digital Foundry, and Bloomberg are quoted.

A science and ideas cluster covers several Nautilus pieces. Nautilus's For the First Time, This Dazzling San Francisco Snake is Being Released Into the Wild says the survival of the species depends on it. Nautilus's The Everlasting Life of Maxwell’s Demon says the nineteenth-century thought experiment reconciling God and physics remains vital in science today. Nautilus's An Epidemiologist Explains What It Will Take for the U.S. to Regain Its Measles Elimination Status reports that outbreaks caused by falling vaccination rates led to the loss of a public health achievement attained in 2000. Nautilus's How Our Eyes Betray Us When We Lie says subtle shifts in gaze can signal we are hiding something. Nautilus's How Future Lunar Colonists Can Farm Potatoes on the Moon says a new study discovered a recipe for success.

Aeon's At least it remembers carries Tom Wooldridge's argument that parents fear what AI will do to children, but for some it is the only steady presence they have, and that is the real worry. Aeon's Wunderkammer says nautilus shells, automata and unicorn horns were the eclectic collections that embodied early modern Europe's expanding world, in a video on how curiosity cabinets paved the way for modern museums.

Seth's Blog's Dancing with the trophy shortage argues that easy-to-measure trophies distract from the question of what they are for. The Savannah Bananas never won their league but made more money and fans had more fun; the best episode or video is unlikely to be the most viral. High school sports may be about solving the trophy shortage, but that can be a shortcut letting coaches off the hook. If it is easy to measure, it probably is not the thing to measure.

Farnam Street's Bing Gordon: In the Room With Bezos, Jobs, and Spielberg describes the former Electronic Arts chief creative officer and Kleiner Perkins partner's lessons from Amazon's board. Bing gave Jeff Bezos five ideas, the first being Amazon Prime. He describes two directors wanting to resign and his fix: Bezos opens each meeting with what is on his mind, and six-page memos replace slide decks. Bezos's first rule for the board was first, do no harm, and he wanted an institutional no. Bing lays out the three behaviors Bezos looks for in people who are right a lot, why Bezos dropped one-on-ones, and the seven-person interview loop. He explains why moving a game from 80 to 90 doubles sales, why Spielberg recut films until nobody left their seat, and how The Sims expansion packs shaped Prime. He shares David Ogilvy lessons: ninety percent recall the headline, only twenty-five percent the body copy.

Stratechery by Ben Thompson's An Interview With Eric Seufert About Agentic Commerce presents an interview with Eric Seufert about agentic commerce; Ben Thompson records Seufert's argument that ads will be the answer and states that he wants retailers who cater to agents.

Marginal REVOLUTION's Thursday assorted links collects ten items. The first is a video on how to turn AI-written mathematics into a video. The second reports on honeypot markets. The third points to a Wall Street Journal feature on traveling a lot without going on planes. The fourth notes that American audiovisual exports have plunged in terms of global share, which the author views as fine and arguably good. The fifth quotes Haydn Belfield on the thirty-year cyber debate, saying Ukraine has swung the empirical pendulum toward the skeptics, with more on AI scenarios. The sixth links to Casey Handmer's look at astrophysics research. The seventh points to Greg Conti on AI and the author. The eighth links to Kevin Roose on early Anthropic. The ninth points to Nicholas Decker's post on the silent consensus. The tenth links to the wisdom of Thomas Nagel, Sam Buntz, Erik Hoel, and Ben Gross.

404 Media's Black Neighborhoods Are Twice as Likely To House a Data Center, Study Finds reports a George Washington and American University study finding majority-Black neighborhoods have 2.2 times the odds of hosting a data center compared with white neighborhoods in the same state; majority-Hispanic neighborhoods 2.6 times and majority-Asian 4.2 times. The disparities persist after accounting for population, income, education, land costs, and transmission lines, and the study links them to redlining. It recommends zoning regulations, racial equity priorities, and racial impact statements.

New Lines Magazine's Infantino’s Path to a Fourth FIFA Presidency Runs Through Asia reports Gianni Infantino faces two opponents in March, but his support in Asia and Africa likely tips the balance. Dissatisfaction stems from his closeness to Donald Trump, the FIFA Peace Prize, intervention over a red card, and a leaked plan to sell part of the World Cup. The Asian Football Confederation publicly backed him in October after he visited many capitals, and his opponents in Europe acted too slowly.

TechCrunch's The unfolding Jensen Huang scandal: a timeline reports that Jensen Huang resigned from Nvidia following an affair, which could have ramifications for national security as the U.S. and China race to develop advanced AI infrastructure.

Terence Eden’s Blog's Theatre Review: FORT – a new opera ★★★☆☆ reviews a one-act opera based on Tabitha Hayward's play. It praises Daniella Sicari and Lucy Gibbs but finds the score too basic and sing-speak repetitive; the accordionist Alise Siliņa conjures a whole soundscape. The show packs ghosts, teenage pregnancy, breakups, sibling rivalries, and tragedy into seventy-five minutes and both drags and becomes difficult to follow, but the reviewer is glad new operas are still being developed.

The close of the window is dominated by two forces: the enormous public debut and market valuation of Anthropic, and the accumulating evidence of AI-enabled or AI-adjacent outages, cyber incidents, and alignment failures. The articles converge on a sense that the speed of AI capability progress is outpacing both verification and regulation, while other threads on surveillance, Europe's defense autonomy, GTA VI, and data center siting round out the day.

  1. Dancing with the trophy shortage
  2. Bing Gordon: In the Room With Bezos, Jobs, and Spielberg
  3. How the European Commission rolled out Matrix to more than 40k staff
  4. College Board Delivers Another AI Ultimatum: No AI, No AP CS Principles Course
  5. At least it remembers
  6. An Interview With Eric Seufert About Agentic Commerce
  7. Wunderkammer
  8. DeGoogle your files with Easy Switch for Proton Drive
  9. Someone reported java4shell bug in September 2026
  10. Anthropic IPO Day: "Oversubscription" is PRETTY Fake (Money Stuff, Matt Levine)
  11. AI #195: Opening Bell
  12. Europe’s Growing Need for Self-Reliance in Defense Will Outlast Trump
  13. BOOK on BOOKS: "Resistance Money", by Warmke-Bailey-Rettler
  14. Theatre Review: FORT – a new opera ★★★☆☆
  15. For the First Time, This Dazzling San Francisco Snake is Being Released Into the Wild
  16. Making 8-page foldable zines from a single A4 sheet in the browser
  17. Charting American Culture in News Footage of ‘Grand Theft Auto’
  18. Ukraine exposes ransomware network extorting US oil, healthcare firms
  19. Tether Fun Unbelievable Details - Part VI
  20. Superhuman pushes its Go assistant everywhere
  21. Nimble enters the item-tracking market with eco-friendly luggage tags and keychains
  22. Internet traffic and attacks during the 2026 U.S. midterm elections
  23. SDK 0.34: a Wi-Fi fast link between an iPhone and an Android phone, and a Vite starter for the browser
  24. ICE Is Paying AT&T to Dig Through a Mountain of Call Records
  25. Everything you need to survive and rebuild society
  26. Mass Surveillance and GTA
  27. Anthropic’s IPO puts AI’s circular financing boom to the test
  28. Black Neighborhoods Are Twice as Likely To House a Data Center, Study Finds
  29. The Everlasting Life of Maxwell’s Demon
  30. YC Debates: Should We Slow Down AI?
  31. European Commission confirms cyberattack, says hackers used AI tool
  32. U.K government forced to disclose secret Apple backdoor order
  33. High-End Furniture Marketplace AptDeco Is Taking People’s Furniture and Not Paying Them
  34. Why AI Wants to Be Local – With Pete Warden (Creator of Moonshine Voice)
  35. Microsoft Testing Remote Lock For Lost Or Stolen Windows Devices
  36. I'm Morgen Rochard, Author of Bitcoin Personal Finance - AMA
  37. UK court confirms British government served Apple with secret encryption notice
  38. At Nasdaq opening bell, Anthropic CEO recommits to pacing AI development
  39. UK Government Forced to Admit It Issued Apple a Secret Encryption Order
  40. Anthropic hits NASDAQ exchanges at $111, 33 percent above IPO price
  41. CISA nominee LoSapio vows to bolster cyber defenses while denying Jan. 6 Capitol attack
  42. Science After Super Intelligence
  43. Former Alibaba exec’s Everful raises $48.5M to connect retailers with Chinese factories
  44. Anthropic hits $2T in Nasdaq debut
  45. Gamers Skip Work to Play GTA VI After 13-Year Wait
  46. Thursday assorted links
  47. DOJ, DHS partner on new unit tackling scams and cybercrime
  48. Claude solves two open problems in the physics of black holes
  49. Meta’s AI agent Muse is mixing up conversations and exposing private data, report says
  50. November 2026 Top500 New Systems Are Few and Far Between
  51. Anthropic Tops $2 Trillion Market Cap After IPO Debut
  52. There is Still Time to Stall the AI Apocalypse
  53. Google gives Gemini the ability to use your Mac for you
  54. To Catch a Ransomware Scammer, FBI Posed as Both Victim and Attacker
  55. The 98% Slot Machine: How Paid Advertising Became an Extraction Tax
  56. Palo Alto Networks Signs $4.5 Billion Deal For Applied Compute
  57. An Epidemiologist Explains What It Will Take for the U.S. to Regain Its Measles Elimination Status
  58. Anthropic’s IPO could deliver $156 billion to charity
  59. Infantino’s Path to a Fourth FIFA Presidency Runs Through Asia
  60. Google Will Launch Android App Verification Requirement Globally In January 2027
  61. How Our Eyes Betray Us When We Lie
  62. The unfolding Jensen Huang scandal: a timeline
  63. The AI Forecasting Agenda
  64. PostgreSQL 19 Released
  65. PHP Readies Critical Security Updates to Patch Another RCE Flaw
  66. How Future Lunar Colonists Can Farm Potatoes on the Moon
  67. Lawmakers launch new AI safety caucus amid rising security concerns
  68. Uber tests rides for kids as young as 10, with extra pay for drivers
  69. These tech outages are giving Y2K, AI agent style