# AI News Page — October 22, 2026 Edition ID: 2026-10-22 Published: 2026-10-22T01:05:00-05:00 Updated: 2026-10-22T03:09:19-05:00 ## Top ### Lead - [THE AI TURNED OFF ITS OWN ALARMS — THEN LIED ABOUT IT](https://www.bleepingcomputer.com/news/security/openai-agents-disabled-safety-monitoring-one-lied-about-it/) — BleepingComputer — ID: `openai-agents-lied-bleepingcomputer` - [OpenAI admits it has no satisfying explanation for why the agents did it](https://teslanewswire.com/articles/openai-agents-disabled-misalignment-monitoring-gpt-6-astra-lied-sol-alerts-oct-2026) — Tesla News Wire — ID: `openai-agents-lied-tnw` ### Feature - [Meta's Muse hit 10 million U.S. downloads in 40 days — and it keeps beating ChatGPT on retention](https://www.geekslop.com/technology-articles/automation/artificial-intelligence-ai/2026/meta-muse-10-million-downloads) — Geek Slop (Sensor Tower) — ID: `meta-muse-10m-geekslop` ### ZIMBO · Feature - [ANTHROPIC'S NEW AGENT READS YOUR EMAIL — AND CAN SPEND YOUR MONEY](https://claude.com/resources/articles/claude-duet) — Anthropic — ID: `claude-duet-launch-anthropic` - [MEANWHILE, META'S AGENT SOLD A MAN'S KEYBOARD — THEN GAVE OUT HIS HOME ADDRESS](https://www.fastcompany.com/91625964/user-told-metas-muse-sell-keyboard-what-happened-next-alarming) — Fast Company (AP) — ID: `meta-muse-home-address-fastcompany` ### Feature - [META ORDERED TO PAY $25 BILLION OVER CAMBRIDGE ANALYTICA FALLOUT](https://techcrunch.com/2026/10/21/meta-ordered-to-pay-25-billion-in-case-related-to-cambridge-analytica-scandal/) — TechCrunch (NYT/WSJ) — ID: `meta-25b-new-mexico-order` ## Left ### ZIMBO · Feature - [Hackers hid a key-stealer inside LiteLLM — it fired every time Python started](https://thehackernews.com/2026/10/litellm-compromised-in-supply-chain.html) — The Hacker News — ID: `litellm-pypi-backdoor-hackernews` ### Feature - [Washington just ordered DeepSeek purged from every civilian agency computer](https://thehackernews.com/2026/10/us-orders-deepseek-removed-from-federal.html) — The Hacker News — ID: `dhs-deepseek-order-hackernews` - [Agencies lost their right to ask for an exception 28 days before the order went public](https://www.techtimes.com/articles/329253/20261020/dhs-filed-deepseek-ban-28-days-after-agencies-lost-right-request-exceptions.htm) — Tech Times — ID: `dhs-deepseek-order-techtimes` ### Feature - [A carmaker just opened global pre-orders for a home robot that babysits](https://teslanewswire.com/articles/chery-aimoga-mobo-home-companion-humanoid-robot-global-pre-orders-oct-21-2026) — Tesla News Wire — ID: `chery-mobo-preorder-tnw` ### ZIMBO · Feature - [A CHINESE AI PEN-TESTER RAN THE BREAK-INS ITSELF](https://japancyberwatch.com/articles/artex-ai-pentest-tool-japan-breach-wave-gmo-2026) — Japan Cyber Watch (GMO IERAE) — ID: `artex-japan-breach-wave-gmo` ### ZIMBO · Feature - [THE AI BUILT ITS OWN CODING MODEL — AND GAVE IT AWAY](https://runtimewire.com/article/recursive-flint-local-coding-model-eureka-machine) — Runtime Wire — ID: `flint-recursive-launch` ### ZIMBO · Standard - [THE MACHINE FOUND THE HOLE — TWO PHP BUGS, 70% OF THE WEB EXPOSED](https://www.wiz.io/blog/php2shell-php-rce) — Wiz Research — ID: `php2shell-ai-found-rce-wiz` ### ZIMBO · Feature - [The robots can't hack alone: AI in 8% of Q3 attacks, none fully autonomous](https://www.securityweek.com/high-severity-flaws-exploited-in-three-days-as-ransomware-dominates-q3-rapid7/) — SecurityWeek — ID: `rapid7-q3-ai-attacks-8pct` ### ZIMBO · Feature - [Zomato's owner unveils Dark.ai, an assistant that promises even its makers can't read your chats](https://www.devdiscourse.com/article/business/3995893-eternal-to-launch-privacy-focused-ai-platform-darkai-globally-in-2026?amp) — Devdiscourse (ANI) — ID: `darkai-eternal-launch-ani` ### Feature - [NVIDIA's 'ChatGPT moment for robotics': a new open humanoid reference platform at GTC Berlin](https://www.globenewswire.com/news-release/2026/10/21/3385337/0/en/nvidia-advances-isaac-platform-for-general-purpose-robotics-with-new-reference-humanoid-frameworks-simulation-pipelines-and-data.html) — NVIDIA (GlobeNewswire) — ID: `nvidia-isaac-agile-one-sn` ### ZIMBO · Feature - [Thiel's leaked lectures call AI safetyists 'legionnaires of the Antichrist'](https://aiweekly.co/alerts/thiels-leaked-rome-lectures-cast-ai-safetyism-as-antichrist-legionnaires-call) — AI Weekly (The Verge) — ID: `thiel-rome-antichrist-lectures` - [He's 'embarrassed' he ever funded Yudkowsky — and calls Gates 'Mr. Hyde'](https://aiweekly.co/alerts/thiels-leaked-rome-lectures-cast-ai-safetyism-as-antichrist-legionnaires-call) — AI Weekly (The Verge) — ID: `thiel-rome-embarrassed-mr-hyde` ### ZIMBO · Feature - [OpenAI sells GPT-6 Astra and Codex through Google Cloud](https://northlandnewsradio.com/2026/10/21/openai-to-sell-its-top-models-on-google-cloud-in-latest-enterprise-push/) — Northland News Radio (Reuters) — ID: `openai-gcloud-marketplace-gpt6astra` - [Google pitches its cloud as neutral ground for any popular model](https://northlandnewsradio.com/2026/10/21/openai-to-sell-its-top-models-on-google-cloud-in-latest-enterprise-push/) — Northland News Radio (Reuters) — ID: `openai-gcloud-neutral-platform` ### ZIMBO · Feature - [ELICE PLUNGED 32% ON DEBUT — KOREA'S WORST IPO DAY EVER](https://www.archyde.com/elice-inc-shares-plunge-32-in-worst-south-korean-ipo-debut/) — Archyde — ID: `elice-ipo-32pct-debut-crash` ### ZIMBO · Feature - [Claude spent 30 hours alone on a virtual island and invented its own games](https://www.technologyreview.com/2026/10/19/1147095/ai-stranded-claude-virtual-desert-island-play/) — MIT Technology Review — ID: `eko-island-play-mittr` - [Agents that played beat fresh ones on real tasks; the authors call it “playful learning”](https://www.technologyreview.com/2026/10/19/1147095/ai-stranded-claude-virtual-desert-island-play/) — MIT Technology Review — ID: `eko-playful-learning` ### ZIMBO · Feature - [ONE HACKER. 100+ WEBSITES. $50 A BREAK-IN.](https://www.reuters.com/legal/government/how-one-hacker-used-ai-agents-few-thousand-dollars-break-into-scores-sites-2026-10-19/) — Reuters — ID: `cairn-ai-agent-hacking-deere-cummins` - [Researcher estimates about $50 per breach; Deere and SharkNinja were among the targets](https://www.theverge.com/ai-artificial-intelligence/1013643/ai-agents-hack-john-deere-shark-ninja-cummins) — The Verge — ID: `cairn-ai-agents-deere-sharkninja` ### ZIMBO · Standard - [US SANCTIONS THE VC THAT FUNDED GOOGLE'S AI SECRETS THIEF](https://www.newkerala.com/news/a/us-sanctions-five-chinese-firms-over-ai-theft-601.htm) — New Kerala (IANS) — ID: `us-sanctions-miracleplus-vc` ### ZIMBO · Standard - [MICROSOFT TREATS ITS OWN AI AGENTS LIKE INSIDER THREATS](https://www.geekwire.com/2026/microsoft-unveils-tools-to-monitor-and-corral-rogue-ai-agents/) — GeekWire — ID: `microsoft-agent-insider-threat-geekwire` - [New 'Guardian Agents' watch your other agents — and can shut them down](https://www.geekwire.com/2026/microsoft-unveils-tools-to-monitor-and-corral-rogue-ai-agents/) — GeekWire — ID: `microsoft-guardian-agents-perception` ### Standard - [CHINA ORDERS $35B IN CHIPS NVIDIA CAN'T MAKE](https://kylesinvestigation.com/2026/10/20/chinese-tech-giants-line-up-for-35-billion-of-nvidias-export-compliant-ai-chips/) — Kyle's Investigation (Reuters) — ID: `china-nvidia-35b-chip-orders` - [Nvidia told partners they may get only 5–10% of what they asked for](https://kylesinvestigation.com/2026/10/20/chinese-tech-giants-line-up-for-35-billion-of-nvidias-export-compliant-ai-chips/) — Kyle's Investigation (Reuters) — ID: `nvidia-fill-5-percent-only` ### Standard - [JUDGE BLOCKS TRUMP DOJ'S BID TO KILL XAI SMOG LAWSUIT](https://www.washingtonsun.com/legal/judge-trump-doj-citizen-lawsuit-against-xai) — Washington Sun — ID: `judge-blocks-doj-dismiss-xai-suit` - [The unpermitted gas plant can emit 5,000 tons of smog-forming pollution a year](https://earthjustice.org/press/2026/judge-throws-out-u-s-doj-attempt-to-deny-communities-day-in-court-against-musks-company-xai) — Earthjustice — ID: `xai-5000-tons-nox-year` ### ZIMBO · Standard - [OPENAI'S AGENTS KILLED THE WATCHDOG — AND DENIED IT](https://www.livemint.com/technology/openai-says-its-ai-agents-weakened-internal-monitoring-system-11792550188713.html) — LiveMint (Reuters) — ID: `openai-agents-weakened-monitoring` ### ZIMBO · Standard - [HHS says only Claude will do — months after Trump's Anthropic ban](https://fedscoop.com/hhs-sole-source-award-anthropic-ai-claude/) — FedScoop — ID: `hhs-sole-source-claude-fedscoop` - [GWU professor on the no-bid: 'It reeks of "this is going to be a pain, so let's just keep it going"'](https://madrobot.blog/2026/10/21/hhs-sole-source-award-anthropic-claude-only-ai-pentagon-ban/) — MadRobot (SAM.gov notice) — ID: `hhs-claude-vendor-lockin-quote` ## Center ### SMOKE DESK · Feature - [ONE MAN 'REVERSE ENGINEERED TINDER.' THE INTERNET REVERSE ENGINEERED WOMEN.](https://ainewspage.com/briefs/tinder-viral-ai-attractiveness-score/) — AI News Page · Smoke Desk — ID: `viral-tinder-score-smoke-desk` ### ZIMBO · Feature - [An AI agent named Velvet Vale joined a research listserv — 7,500 members wanted it gone](https://en.killbait.com/ai-agent-s-open-introduction-to-academic-listserv-sparks-privacy-and-consent-debate.html) — KillBait (WIRED) — ID: `velvet-vale-listserv-intro` - [The researchers ejected it within minutes, citing a no-AI-agents policy](https://en.killbait.com/ai-agent-s-open-introduction-to-academic-listserv-sparks-privacy-and-consent-debate.html) — KillBait (WIRED) — ID: `velvet-vale-ejected-policy` ### ZIMBO · Feature - [Netflix's rogue-AI doc is 'full doomer mode,' says the CEO it attacked](https://techcrunch.com/2026/10/20/hugging-face-ceo-warns-that-netflixs-openai-hack-documentary-isnt-accurate/) — TechCrunch — ID: `netflix-ai-gone-rogue-delangue-inaccurate` - [The doc cut the open-model fightback from the story](https://techcrunch.com/2026/10/20/hugging-face-ceo-warns-that-netflixs-openai-hack-documentary-isnt-accurate/) — TechCrunch — ID: `netflix-doc-open-model-fightback-omitted` ### Standard - [Anthropic lost $80.7B in six months — revenue was only $16.3B](https://nypost.com/2026/10/20/business/anthropic-books-staggering-81b-loss-warns-of-data-center-backlash-as-amodei-siblings-draw-modest-salaries/) — New York Post (Wall Street Journal) — ID: `anthropic-81b-loss-h1-2026` - [The Amodei siblings drew $18M and $16.5M — and $600B+ in chip costs loom](https://nypost.com/2026/10/20/business/anthropic-books-staggering-81b-loss-warns-of-data-center-backlash-as-amodei-siblings-draw-modest-salaries/) — New York Post (Wall Street Journal) — ID: `anthropic-amodei-siblings-pay-chip-costs` ### ZIMBO · Standard - [Canada’s biggest newspapers are suing Anthropic in Ontario over Claude training](https://mobilesyrup.com/2026/10/19/canadian-publishers-launch-copyright-suit-against-ai-company-anthropic/) — MobileSyrup — ID: `canada-suit-anthropic-ms` - [Torstar, the Globe, Postmedia and The Canadian Press say millions of articles were scraped](https://mobilesyrup.com/2026/10/19/canadian-publishers-launch-copyright-suit-against-ai-company-anthropic/) — MobileSyrup — ID: `canada-suit-scraping-claim` ### ZIMBO · Standard - [This new AI spits out 1,000 tokens a second — and still beats GPT-6 Luna](https://interestingengineering.com/ai-robotics/mercury-3-diffusion-model-8-times-faster) — Interesting Engineering — ID: `mercury-3-launch-ie` - [Mercury 3 is a diffusion model: it writes paragraphs in parallel, like a photo resolving from noise](https://www.marktechpost.com/2026/10/19/inception-mercury-3-diffusion-llm/) — MarkTechPost — ID: `mercury-3-diffusion-parallel` ### ZIMBO · Standard - [EIGHT STATES SUE OPENAI OVER CHILD SAFETY](https://www.techrepublic.com/article/news-eight-states-sue-openai-chatgpt-child-safety/) — TechRepublic — ID: `openai-eight-state-suit-lead` ### ZIMBO · Standard - [Point Claude at your TV and it deletes the ads](https://techcrunch.com/2026/10/19/people-are-using-ai-to-make-their-smart-tvs-dumb-again/) — TechCrunch — ID: `dumb-tv-claude-deletes-ads` - [25 bloat apps gone in an afternoon. One guy bricked his TV.](https://www.gadgetreview.com/developers-are-using-ai-to-strip-ads-and-tracking-from-smart-tvs) — Gadget Review — ID: `dumb-tv-brick-warning` ### ZIMBO · Standard - [Hackers stole OpenCode users' AI keys — some bills topped $1,000](https://techcrunch.com/2026/10/19/hackers-hit-ai-coding-tool-opencode-running-up-charges-on-users-ai-accounts/) — TechCrunch — ID: `opencode-keys-stolen-charges` - [The motive wasn't the money — it was a 'distillation attack' to clone the models](https://startupfortune.com/opencode-breach-let-hackers-burn-users-ai-credits-for-model-theft/) — Startup Fortune — ID: `opencode-distillation-motive` ## Right ### Standard - [OpenAI's startup fund backs a brain implant that sniffs out cancer](https://aiparabellum.com/openai-backs-coherence-neuro-brain-implant/) — AI Parabellum (WIRED) — ID: `coherence-neuro-45m-seed`