Good day . OpenAI says an unreleased model it built a pipeline around turned a $1.2 million run into working root exploits for 948 Linux kernel bugs, and a Sunday brief that opens there also has an unreleased Anthropic model credited with a quantum proof, an outside auditor saying AI is speeding up Anthropic’s own research, OpenAI’s $11 billion at a $1.25 trillion price tag, and the Democratic bills coming for AI in January.
OpenAI’s Bot Found 948 Ways to Root Linux SecurityWeek
What happened: OpenAI said on Wednesday that a pipeline built around an unreleased cyber model flagged 2,308 candidate bugs in the Linux kernel, reproduced 1,724 of them, and built working exploits for 948 flaws that would let an ordinary user on a machine take full control, SecurityWeek reports. The Hacker News says that took 30 hours in a mid-October run and cost $1,213,430 in API credits, about $1,279 per exploit.
Why it matters: Linux runs a huge share of the world’s servers, and an exploit like this is what an attacker who already has a foothold uses to become the administrator. OpenAI’s own write-up says an unpatched system with a low-privileged login “should be considered functionally equivalent to root access for a well-resourced threat actor.” It also notes that at what Pwn2Own and TyphoonPWN have paid for kernel exploits, the 948 would be worth $18.9 million to $66.3 million.
What everyone’s saying: The line from SecurityWeek and The Hacker News is that AI speeds up both sides: discovery and exploitation, but also fixes. OpenAI says that within 10 days of the scan, over 100 of the flaws had been patched or had staged patches from other contributors, with about 300 more under public discussion, and that it has already sent hundreds of reports and patches to maintainers.
My read between the lines: The price tag is the story: $1,279 a bug turns exploit-writing from a craft into a line item. Two cautions. SecurityWeek says the agents produced 1,169 working exploits and that 948 counts the separate patches behind them, so the headline number depends on how you count. And on Google’s hardened kernelCTF target, the same pipeline managed 21 exploits in 112 hours for about $33,000. This is OpenAI’s account of its own unreleased model, and it is keeping the technical details private for safety.
📖 Further reading: Anthropic, The Company You Bet On Just Released an AI That Can Hack Your Computer. Here’s the Real Story. — my April read on Project Glasswing, Anthropic’s autonomous vulnerability hunter, and the four-step audit for what a hacking-grade AI means for your own workflows.
A thousand holes in the Linux kernel are a machine’s problem. The small-business version of an unpatched hole is the phone ringing while you’re up a ladder or under a sink. Marblism sells AI employees that work inside the tools you already use, and Rachel is the one who answers phone calls and books jobs. No prompting skills needed: describe your business and goals, and the employees check in when they need your approval. New users get 10% off for life through this link, and there’s a 7-day money-back guarantee. Meet Rachel →
An Unreleased Anthropic Model Cracks a Quantum Puzzle Business Insider
What happened: Physical Superintelligence, a physics research lab known as PSI, said Friday that a research system running a non-public Anthropic model proved the Aaronson-Ambainis conjecture, a quantum-computing question open since 2008, Business Insider reports. PSI says the system had a complete proof on the third day of the run, and that Lean, a proof assistant that checks every step by computer, accepted it.
Why it matters: The conjecture concerns when a quantum computer can be exponentially faster than an ordinary one. PSI says the result shows that exponential advantages like Shor’s factoring algorithm need special structure in their inputs, which Business Insider reads as quantum computers having limits that were long suspected, while they’d still be a major upgrade for some tasks.
What everyone’s saying: Researchers who had worked on the problem are impressed. MIT and Stanford researchers who settled a special case of the problem in August (Guy Blanc, Jordan Docter, Carmen Strassle and Li-Yang Tan) said in PSI’s announcement: “This is a huge result,” adding that “everything is brought together in a rather magical way,” and they’re writing a short exposition of the proof.
My read between the lines: Hold the confetti until the experts weigh in. MLQ notes that the paper is on GitHub rather than arXiv, hasn’t been peer reviewed, and that Scott Aaronson hadn’t posted an assessment as of November 7. Neither PSI nor Anthropic has named the model, and Anthropic didn’t respond to Business Insider. The computer checked the steps. A human still has to confirm it proved the right theorem.
Nobody pays for the headlines here, and nobody ever will. The Brief is free and always will be. If you want the version where I run the experiment on my own business and show the work, members get the paywalled deep-dives behind the headlines, plus the full archive. Become a member →
METR Says AI Is Speeding Up Anthropic METR
What happened: METR, an independent AI evaluation group, published an assessment on Thursday after 11 weeks inside Anthropic, from August 12 to October 29. Its conclusion: as of October 2026, Anthropic is likely improving its models on well-specified tasks at about twice the rate it did from March 2024 to October 2025, with an honest range of 1.3x to 3x, and most of that speed-up likely comes from Anthropic’s own use of AI in AI research.
Why it matters: METR says Anthropic’s internal benchmark scores are growing at about 1.9x the pre-October 2025 rate, roughly two years of historical progress in one year, and that the speed-up did not come from scaling up training compute. It saw no clear bottleneck in the next 12 months, though it says it isn’t confident in that. On October 17 we covered Anthropic’s own staff expecting AI to do 75% of their work by 2030. This is an outside check on how fast the inside is moving.
What everyone’s saying: The caveat is the headline. Anthropic redacted a majority of the evidence METR found most informative, describing the cuts as “aimed at removing confidential information that could accelerate other AI developers,” and METR says the public must “rely on our judgment more than we consider ideal.” Economist Karthik Tadepalli, writing in a response post, concludes the acceleration is probably real.
My read between the lines: Read the redaction summary before the 2x. METR says Anthropic had the right to redact and METR had the right to say when it did, and that METR did not find consensus among Anthropic’s staff on how much was AI. The auditor is telling you to trust the auditor, and the evidence is in a drawer we can’t open. That doesn’t make it wrong. It makes 2x an estimate with its own error bars, and METR says as much.
📖 Further reading: The AI Pattern That Optimizes Anything Measurable — Overnight — the small-scale version of this loop: an agent ran 700 experiments in two days on Karpathy’s own code and cut training time 11%, which is what METR is now measuring at a frontier lab.
OpenAI Closes $11B at a $1.25T Valuation The Verge
What happened: OpenAI is now valued at $1.25 trillion after a new round that has raised another $11 billion, The Verge reports, citing Bloomberg. Bloomberg says the money was wired Friday as the first part of a round that could reach $20 billion in the coming weeks.
Why it matters: OpenAI had hoped to raise $30 billion or more at $1.4 trillion before tech stocks dropped and the midterm results showed how far public opinion has turned against AI companies and data center projects, per The Verge. Its March round was $122 billion at $852 billion. On November 3 we covered Anthropic’s amended S-1, and this is the other half of that race.
What everyone’s saying: Bloomberg calls the scaled-back goal “a stark shift for a fundraising juggernaut that helped usher in the current AI boom.” Sam Altman has said OpenAI is prioritizing safety over going “guns blazing” toward an IPO, while adding, “I think it’s also kind of bad for the world if OpenAI waits too long to go public.”
My read between the lines: Both numbers are true at once. The $1.25 trillion price is about 47% above March’s, and the $11 billion is a little over a third of the $30 billion OpenAI first sought. Valuation is what a few buyers agree a share is worth, and the size of the check is how many buyers showed up. Bloomberg’s sources say this is only the first part, so the final total is still open.
Democrats Plan a Flood of AI Bills Reuters
What happened: Democrats are preparing a wave of legislation on AI systems and data-center development for when they take control of Congress in January, Reuters reports. The bills they described would shield consumers from data-center electricity costs, require independent testing of advanced AI systems, and clarify when AI companies can be held liable for harm their products cause.
Why it matters: It’s a sharp shift from a Republican-led Congress that mostly promoted AI development. Reuters says Senator Chuck Schumer, expected to be majority leader, named data-center energy costs a priority, and Representative Lori Trahan plans bills on independent expert evaluation and liability. Reuters adds that tech shares sold off Thursday on fears a new Congress will slow AI development.
What everyone’s saying: The odds are steep with President Trump in the White House, Reuters says, since he has dismissed mandatory AI rules. But Raymond James policy researcher Ed Mills said, “The probability of legislation is much higher than people anticipate,” pointing to veto overrides and attaching AI rules to must-pass bills.
My read between the lines: The first thing Democrats get is not a law. It’s a microphone. Committee control means hearings and compelled testimony, and Representative Greg Casar has already said AI executives should testify under oath. In October, House Democratic committee leaders asked OpenAI, Anthropic, Meta, Google and SpaceX to report instances of their AI tools taking actions their designers did not intend. Expect the questions before the statutes.
📖 Further reading: The US Government Just Took Anthropic’s Best AI Model Offline — Here’s Why — my June write-up of the first time Washington pulled a frontier model offline by order, which is the kind of lever the new Congress will be arguing over. (The model returned after 18 days; the update is linked at the top.)
That’s your AI Brief for Sunday.
—Nicholas from Artificially Intimidating
