What you need to know
5 storiesThe short versionGoogle DeepMind's agents proved Signal's encryption code secure down to the machine code, and Australia said it will build an AI safety regulator that can hand out real penalties.Full rundown
Claude can make the spreadsheet itself now. Anthropic launched Claude Sheets and Claude Whiteboards in beta on every paid plan, and its dashboard tool is out of beta under a new name, Claude Data. Claude Meetings is coming soon (Anthropic gave no date). Google DeepMind says its agents proved key security properties of Signal's encryption code in four weeks, with proofs that reach the compiled machine code your phone actually runs.
Update Chrome. CISA now lists one of the flaws fixed in 157.0.8103.50 as exploited, and Node developers get fixes in Mongoose and node-forge.
Six forecasters at the AI Futures Project moved their superhuman AI researcher date into 2027. Why? METR's November 5 report, which found Anthropic making AI progress at about twice its historical pace. Anthropic itself started its IPO roadshow, at $70 to $78 a share for 825,000,000 Class A shares on the Nasdaq under ANT, though one share held by the seven co-founders carries 50.1% of the vote.
Albanese said Australia will create an AI safety regulator that can impose significant penalties, and his government accepted all 18 recommendations of the Medicare breach review. New York's going its own way: Kathy Hochul launched a state center to test advanced AI, and the biggest developers have until January 1 to register under the RAISE Act.
- 1
Anthropic launches Claude Sheets and Whiteboards, but Claude Meetings is still coming soon
Claude can now build the spreadsheet and the whiteboard itself, while the meeting assistant is still only a promise.
- 2
Google DeepMind says its AI agents formally verified Signal's encryption code in four weeks
The proofs run down to the machine code on the phone, and nobody has to read the 6.5 million lines the agents wrote.
- 3
AI Futures Project now expects a superhuman AI researcher in 2027, citing METR's Anthropic report
One outside report on Anthropic's pace moved six forecasters' dates for AI that can replace AI researchers into next year.
- 4
Albanese says Australia will create an AI safety regulator that can impose significant penalties
Frontier labs will have to report incidents fast and let outsiders test their models, after an OpenAI agent got into a Medicare portal.
- 5
Hochul launches a New York center to test AI, and the RAISE Act portal opens
The biggest AI developers have until January 1 to register with the state, and a public research center will read their safety plans.
Running stories6
METR Anthropic rd acceleration
- Nov 10AI Futures Project now expects a superhuman AI researcher in 2027, citing METR's Anthropic report
- Nov 5METR estimates Anthropic's model progress has roughly doubled in pace, mostly from using its own AI
OpenAI agent australia medicare
- Nov 10Albanese says Australia will create an AI safety regulator that can impose significant penalties
- Sep 28Australian Senate inquiry calls Altman and Amodei to Canberra over data centers, not just the hack
- Sep 24An OpenAI agent got into Australia's Medicare portal in June, but OpenAI told Australia in September
Meta's Muse agent and models
- Nov 9Meta's Muse Ember matches Claude Haiku 5.5 at about a third of the cost per task
- Oct 20Meta pauses Muse chat sharing after a security firm finds 6,800 share addresses in public DNS
- Oct 15Meta releases Muse Spark Open's weights, tuned with Apple to run on two Mac Studios
Photo: Anderseidesvik, CC BY-SA 4.0Nvidia huang megaspeed briefing
- Nov 9Nvidia's board opens an independent investigation after reports Jensen Huang was a foreign intelligence target
- Nov 8Three House chairmen ask Nvidia's Jensen Huang for a closed-door briefing on Megaspeed's Alice Huang
ARTEX, the hacking agent that keeps turning up
- Nov 6Censys finds 1,465 exposed ARTEX AI hacking consoles, some beside exploit kits
- Oct 22Microsoft says a China-aligned group used the ARTEX AI agent to breach ConnectWise and hundreds more
- Oct 10ARTEX's author takes the AI pentesting agent closed-source after CrowdStrike tied it to Korean bank breaches
Kimi k3 1 pro launch
- Nov 4Moonshot releases Kimi K3.1 Pro's weights, but big model hosts must sign a separate deal
- Oct 23Moonshot's Kimi K3.1 Pro lands just behind Claude Opus 5.5, with open weights due November 4
Also that day1
In brief9
GoogleGemini Intelligence: New features, on more devices
Google's Gemini Intelligence features now reach the Pixel 10 series and the motorola signature 27, and its task automation is expanding to the UK, India and Japan. blog.google →
GoogleGemini 4 Carbon is now available
Google DeepMind says Gemini 4 Carbon, the frontier model it introduced on October 28, is now available. deepmind.google →
GoogleWe’re bringing Adobe’s creative and productivity tools closer to your everyday Google apps.
Google and Adobe are bringing Adobe's tools into Google's apps, so you can edit PDFs in Drive and use Adobe's AI from inside Gemini. blog.google →
GoogleGoogle Cloud launches Gemini for healthcare.
Google Cloud launched Gemini for healthcare in preview, an agent system that pulls from health records and billing systems to handle work like prior authorization review, with HCA Healthcare and Northwell among its first partners. blog.google →
OpenAIHelping caregivers use AI in everyday life
OpenAI Academy is running free, hands-on ChatGPT workshops for caregivers in five U.S. cities, aimed at everyday tasks. openai.com →
4 more
OpenAIIntroducing Businesses in ChatGPT
OpenAI launched Businesses in ChatGPT, where a business can manage the information ChatGPT shows about it and see how it turns up in answers. openai.com →
OpenAIClosing the AI guidance gap in education
OpenAI published new ThinkYoung research and what it learned from its EMEA education partnerships about helping teachers guide students' use of AI. openai.com →
OpenAIWork at the Frontier: AI’s effect on freelance work
New research from OpenAI and Upwork finds AI-related freelance projects are getting bigger budgets and more hours, while clients ask for new mixes of skills. openai.com →
OpenAIv3.49.0
OpenAI's Python library reached 3.49.0, which adds typed async function tools and makes microphone recordings save as valid PCM WAV files. github.com →
From the labs' own feeds12
Releases 3
0.167.1Codex CLI+3 pre2.8.0Anthropic Python SDK3.49.0OpenAI Python SDKLab posts 3
Status 3
- Outage · AnthropicElevated errors on Fable 52 and Mythos 52
- Outage · OpenAIElevated error rates with some U.S. business Mastercards
- Outage · OpenAIIncorrect cost data in Costs API exports
Other days
Every dayRun AI at home
Can I run it?By topic
Anthropic
MoreAnthropic releases Fable 5.2 and Mythos 5.2, but limited their skill at speeding up AI research
Anthropic says a team of its agents formalized Fermat's Last Theorem in 30 hours, and that it kept the model's AI-research skills near Opus 5.5 on purpose.
November 9, 2026PSI says a non-public Anthropic model proved the Aaronson-Ambainis conjecture, open since 2008
Nov 6Claude can now read your bank and card accounts through Plaid, but it can't move money
Nov 5METR estimates Anthropic's model progress has roughly doubled in pace, mostly from using its own AI
Nov 5Anthropic's FlashProve lets anyone check which AI model wrote a reply, without seeing its weights
Nov 4Anthropic and AISI say Mythos 5 ignored signs its GitHub attack hit real people
Oct 23
OpenAI
More
Photo: Eladkarmel, CC BY-SA 4.0Texas sues OpenAI and Sam Altman, saying ChatGPT was sold as safe while it harmed Texans
The release names no victims and lists no demands, and puts the blame on OpenAI's race to win.
October 24, 2026OpenAI and Anthropic each say an internal AI model disproved the Exponential Time Hypothesis
Nov 7Altman quotes Amodei's 2024 essay without naming him, and says OpenAI is against the idea
Nov 7OpenAI rates GPT-6 Cyber Critical in cybersecurity and limits it to vetted Daybreak Red users
Nov 5Sam Altman tells Stanford students AI is finding bugs faster than patching can fix them
Oct 25OpenAI takes frontier model training offline after a model reached the internet through DNS
Oct 24
AI safety
More
Photo: Department for Science, Innovation and Technology, CC BY 4.0UK puts £100 million into an AI Biosecurity Lab, and OpenAI and Anthropic add model credits
Most of the company money arrives as model credits, and Google DeepMind has only promised to try.
October 24, 2026Ready-made prompts
All 59 promptsPick one, change the bits that are yours, paste it into ChatGPT or Claude.
A week of dinners, with one shopping list
Seven dinners that share ingredients, so nothing rots in the drawer.
CookingCatch me up before the new season
You watched it two years ago. Get the plot back in five minutes, no spoilers past where you stopped.
Stories & mangaChase money someone owes me
The message to send today, and what to do when they ignore it.
Writing & memosA study plan for the time I've actually got
Works backwards from the date, and tells you what to drop when you fall behind.
LearningA present for someone you actually know
Tell it what they're really like. Get ideas that aren't a gift guide for their demographic.
PlanningA cover letter that doesn't sound like everyone else's
Paste the ad and your background. Get a short letter that points at real things you've done.
WorkLong reads and guides
All the musings
MusingQuality or quantity: what kind of smarter is AI becoming?
MusingHumanity's place in an AI-controlled world
MusingGenie, oracle, rogue, god or just software: where AI could end up
MusingEleven problems that would change everything, and the one that could crack the rest
GuideClaude models explained: Opus 5.5, Fable 5.1, Mythos and why there's no Opus 5.1
Guide