AI and security, today.

The news, new AI tools, advances in AI and security, from sources we trust. Refreshed every 30 minutes; each one opens the original.

AI

Company news, launches and the business of AI.

TechCrunch3 Nov, 01:49 UTCApple may try to increase revenue by doing…something with third-party AI apps (opens TechCrunch in a new tab)
TechCrunch3 Nov, 00:35 UTCY Combinator’s Garry Tan uses AI agents so much, he spends up to $200K a year on tokens (opens TechCrunch in a new tab)
TechCrunch2 Nov, 23:13 UTCTrajectory raises $45M to make AI model training accessible to nontechnical experts (opens TechCrunch in a new tab)
TechCrunch2 Nov, 22:14 UTCAnthropic warns that it’s very close to RSI capability, aka self-building AI (opens TechCrunch in a new tab)
TechCrunch2 Nov, 19:25 UTCWeb host Squarespace lays off 19% of its employees (opens TechCrunch in a new tab)
TechCrunch2 Nov, 19:19 UTCNew York Times and other outlets sue web browser Brave over alleged content scraping (opens TechCrunch in a new tab)
The Verge2 Nov, 18:56 UTCAnthropic clamps down on Claude’s AI video game clones (opens The Verge in a new tab)
Ars Technica2 Nov, 18:32 UTCComing in 2027: Apple’s $1,000 home robot for FaceTime and switching on lights (opens Ars Technica in a new tab)
Ars Technica2 Nov, 18:05 UTCClaude’s new rules against game cloning won’t stop the AI copying boom (opens Ars Technica in a new tab)
The Verge2 Nov, 18:01 UTCOpenAI pitches data center power plants as a way to protect nearby areas (opens The Verge in a new tab)
OpenAI2 Nov, 18:00 UTCSelf-publishedRewarding defenders with bigger bug bounties (opens OpenAI in a new tab)
OpenAI2 Nov, 18:00 UTCSelf-publishedPreparing our next security sprint to harden OpenAI (opens OpenAI in a new tab)

New AI tools

New tools and releases, from the companies that make them and an independent developer who tries them.

Simon Willison2 Nov, 23:49 UTCllm 0.37 (opens Simon Willison in a new tab)
Simon Willison2 Nov, 13:31 UTCThe U.S. Census Bureau and “Super Intelligence” (opens Simon Willison in a new tab)
Simon Willison1 Nov, 23:24 UTCIf in doubt, delete the uv.lock file (opens Simon Willison in a new tab)
Simon Willison31 Oct, 23:15 UTCHow We Took Over ChatGPT Accounts (opens Simon Willison in a new tab)
Simon Willison31 Oct, 22:10 UTCdatasette-llm 0.1a9 (opens Simon Willison in a new tab)
Simon Willison30 Oct, 21:59 UTCdbreunig/nesformer (opens Simon Willison in a new tab)
Hugging Face30 Oct, 17:57 UTCSelf-publishednemotron-halloween (opens Hugging Face in a new tab)
Hugging Face29 Oct, 13:37 UTCSelf-publishedGranite Switch 4.1.1 — Update models like you update software (opens Hugging Face in a new tab)
Hugging Face29 Oct, 00:00 UTCSelf-publishedAnnouncing Project ProtoStar: Making Publicly Funded Science AI-Ready (opens Hugging Face in a new tab)
Hugging Face29 Oct, 00:00 UTCSelf-publishedWelcome Le Chonk: Mistral Large 4 on Hugging Face (opens Hugging Face in a new tab)
Hugging Face29 Oct, 00:00 UTCSelf-publishedIntroducing AnySpeechRecognizer and AnyDecisionModel: Open Models for Speech and Decisions on Apple Platforms (opens Hugging Face in a new tab)
Hugging Face27 Oct, 17:16 UTCSelf-publishedBuild specialized multimodal models with LFM3 (opens Hugging Face in a new tab)

Advances in AI

Research and breakthroughs from the labs and the people who study them.

MIT News2 Nov, 05:35 UTCSelf-publishedRegina Barzilay and Jim Collins win Pritzker Prize for AI and Science Research Excellence for work in AI-enabled antibiotic discovery (opens MIT News in a new tab)
MIT News30 Oct, 19:00 UTCSelf-publishedQ&A: Hey AI, who should I vote for this year? (opens MIT News in a new tab)
Google DeepMind30 Oct, 14:18 UTCSelf-publishedSpeeding up planning decisions with AI (opens Google DeepMind in a new tab)
Microsoft Research30 Oct, 13:05 UTCSelf-publishedFrom synthetic enterprise data evaluations to real Microsoft Copilot AI confidence (opens Microsoft Research in a new tab)
Microsoft Research29 Oct, 19:19 UTCSelf-publishedMicrosoft Research Focus – Edition #97 (opens Microsoft Research in a new tab)
Google DeepMind28 Oct, 17:00 UTCSelf-publishedGemini 4 Carbon: Our next frontier model (opens Google DeepMind in a new tab)
MIT News27 Oct, 20:55 UTCSelf-publishedLearner collaborates with MIT peers to develop book of AI insights (opens MIT News in a new tab)
Google Research27 Oct, 17:35 UTCSelf-publishedThe sleep framework: A new way for LLMs to retain knowledge in continual learning (opens Google Research in a new tab)
Google DeepMind27 Oct, 16:01 UTCSelf-publishedGenie 4: A new generation of world simulation (opens Google DeepMind in a new tab)
MIT News23 Oct, 21:00 UTCSelf-publishedAt MIT, Martha Minow speaks on sustaining democracy, both by using and overcoming AI (opens MIT News in a new tab)
Google Research23 Oct, 18:51 UTCSelf-publishedBringing foundational models to the front lines of TB screening (opens Google Research in a new tab)
MIT News23 Oct, 04:00 UTCSelf-publishedQ&A: How AI came to model the “messy” human mind (opens MIT News in a new tab)

Security

Breaches, vulnerabilities and advisories. Stories that involve AI are marked.

NCSC3 Nov, 12:00 UTCInvolves AICyber chief: UK facing AI-powered attacks on critical infrastructure within two years (opens NCSC in a new tab)
NCSC3 Nov, 12:00 UTCArtificial intelligence and cyber intrusion (opens NCSC in a new tab)
Krebs on Security3 Nov, 02:24 UTCTech Firms Dismantle Botnet That Turned 1.3M TV Boxes Into Proxies (opens Krebs on Security in a new tab)
Dark Reading2 Nov, 22:27 UTCOne Compromised Laptop, Billions Lost: The Binance Attack-Path Warning (opens Dark Reading in a new tab)
BleepingComputer2 Nov, 21:40 UTCInvolves AIChinese hackers use ARTEX AI in automated attacks against 200 orgs (opens BleepingComputer in a new tab)
BleepingComputer2 Nov, 19:00 UTCFintech platform Sageworks taken offline after suspected cyberattack (opens BleepingComputer in a new tab)
BleepingComputer2 Nov, 18:16 UTCGoogle disrupts million-device XCPE Android malware botnet, bans TV box maker (opens BleepingComputer in a new tab)
The Hacker News2 Nov, 16:57 UTC⚡ Weekly Recap: Critical CVEs, Linux DDoS Botnet, Device-Code Phishing, Android Spyware and More (opens The Hacker News in a new tab)
The Hacker News2 Nov, 15:50 UTCGoogle Disrupts 1.3 Million-Device Android Botnet, Says TV Box Maker Sold Criminal Access (opens The Hacker News in a new tab)
BleepingComputer2 Nov, 15:03 UTCBTCPay Server warns of hot wallet crypto thefts on outdated instances (opens BleepingComputer in a new tab)
BleepingComputer2 Nov, 15:01 UTCYour Mobile Carrier Stores Your Call Logs for Years (opens BleepingComputer in a new tab)
Dark Reading2 Nov, 15:00 UTCInvolves AIMicrosoft's AI Risk Report Misses This Key Detail (opens Dark Reading in a new tab)

Each headline is written by its source and opens the original. Self-published means an organisation writing about its own products or research, not independent reporting. Last updated 3 Nov, 03:09 UTC. Sources: OpenAI, Google AI, MIT Technology Review, The Verge, TechCrunch, Ars Technica, Simon Willison, GitHub Blog, Hugging Face, Google for Developers, Google DeepMind, Google Research, Microsoft Research, MIT News, Quanta Magazine, NCSC, The Hacker News, BleepingComputer, Krebs on Security, Dark Reading.

Be one of the first to use Pelagos.

A private, connected workspace for turning conversations into coordinated work. We’re opening it to a small group first.