Pulled live from TechCrunch, VentureBeat, The Verge, and MIT Technology Review.
METR reports that leading AI models like Opus 5.5 and GPT-6 Astra are improving rapidly at research tasks but getting harder to measure. Current data can't rule out a dangerous acceleration in AI development, the organization warns, though
This is The Stepback, a weekly newsletter breaking down one essential story from the tech world. For more on the vibe-hacking of everything, follow David Pierce. The Stepback arrives in our subscribers' inboxes on Sunday at 8AM ET. Opt in f
Security researchers at Dragos used two million fake AI requests to uncover serious vulnerabilities in the AI supply chain, finding at least three compromised providers, hundreds of stolen AWS keys, and marketplace resellers who were simply
The Lean and Mathlib community has officially confirmed Levent Alpöge's proof of the Riemann hypothesis as correct. Dozens of independent verifiers checked the 17 million lines of computer code. The solution arrived years earlier than expec
Major US financial institutions are banning third-party bots from logging into your accounts—and warning that any costly bot mistakes are strictly your problem.
OpenAI's GPT-6.2 Astra and Sol give much shorter, more direct answers than their predecessors, but they are also less willing to admit uncertainty. Anthropic's Claude models score well ahead on transparency. The article GPT-6.2 Astra and So
One of the items on the agenda for UN diplomats meeting in Geneva this week was working towards rules regarding the use of autonomous weapons, or so-called "killer robots." Eventually, representatives left the meetings agreeing to "start di
Nvidia researchers argue that language models often don't need their full context windows—which are slow, expensive, and prone to errors—and propose an alternative. In tests, periodic summaries and targeted keyword search usually matched fu
A Buddhist monastery in Vermont called Maple wants to establish itself as Silicon Valley's spiritual authority while creating its own AI god. According to a New Yorker report, leader Soryu Forall's group records every breath its members tak
Today, I’m talking with Hayden Field, The Verge’s senior AI reporter, about the Anthropic IPO for the third and final episode in our series about AI in real life. The AI lab behind Claude went public on the Nasdaq last Thursday. The numbers
Chess.com is ready to go a lot bigger than chess. After launching Gambit, a gambling-free poker site, earlier this year, the company is now adding 10 more games: backgammon, checkers, go, hearts, mahjong, shogi, sudoku, solitaire, xiangqi,
OpenAI has released new prompt templates for GPT-6.2 Astra and Sol meant to fix common model behaviors such as overly brief answers, unnecessary follow-up questions, and excessive testing when coding. The API is changing as well, since edit
AI is forcing companies to patch their systems faster as attackers race to exploit newly discovered vulnerabilities. Welcome to internet outage hell.
Anthropic mathematician Levent Alpöge says he used Claude to prove the Riemann hypothesis, math's most famous open problem for 167 years, with a proof formalized in Lean that spans over 250,000 pages. Just hours later, Stanford researchers
Thinking Machines has released Inkling 2 Mini, an open looped transformer that runs 12 of its 24 layers three times per token. This lets the model reason internally for longer without generating extra text. On Artificial Analysis's Intellig
The Nasdaq 100 has slipped into a bear market after its worst weekly loss since the Covid-19 crisis. Nvidia plunged 14 percent as the SEC reviews its accounting, and Anthropic lost nearly all of its gains a day after its record IPO. AI doub
The blockbuster game makes no secret of its contempt for law enforcement. Real police departments using it to recruit new members and scold the public apparently didn’t get the memo.
Google search is making it even easier to access AI tools on the search page.
Meta is “re-evaluating” 2030 sustainability goals. Will others follow?
Test using prompts with unique, trackable secrets proves that data doesn't stay private.
Amid allegations of multiple affairs, including with a direct report and an alleged Chinese intelligence asset, sources tell WIRED the Nvidia board’s main issue was the former CEO’s “lack of candor.”
President Donald Trump promised to put a "tremendous tax" on companies that build data centers outside of the US, even as states and cities around the country are placing restrictions on new applications to build them. "We're not going to a
A new report from the AI evaluation nonprofit METR finds that some AI capabilities are continuing to improve along an exponential trajectory—but it also indicates that models aren’t quite as good at tasks that require judgment as they are a
Anthropic went public on November 19th, 2026. | Photo: Getty Images Anthropic mathematician Levent Alpöge claims he has proved the Riemann hypothesis, one of mathematics's most notoriously difficult problems, with the help of Claude, crac
There was a dayslong Outlook outage. A nationwide ground stop for only the second time in 25 years. Banks seemed to suddenly stop working. Even iCloud couldn't stay up and running. This week, it felt like the technological world simply coll
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. What we learned when our Roundtables explored the quest to make minds immortal To preserve our sense o
Signal users on iOS will see this option to toggle blocking screenshots and preventing Siri AI from scanning on-screen content. | Image: Signal Signal is restricting the ability to take screenshots on Android and iOS devices in an effort
Meta’s AI assistant can track and unfollow Instagram users who don’t follow you back—and some people are taking full advantage.
The Jensen Huang scandal is the latest fuel for dating anxieties among newly rich and powerful tech workers.
OpenAI’s layered approach to adversarial distillation combines model training, safeguards, detection, and enforcement to protect against capability extraction.
Henrijs Dandzbergs dropped out of high school to join startup accelerator Y Combinator. With a new breed of missile, he wants to make killing as efficient as possible for the US and its military allies.
Hebbia tested GPT-6 Astra for financial research. Its prerelease evaluations found gains in source coverage, citation relevance, and presentation quality.
MedChemBench evaluates how well AI models interpret molecular structures, reason from experimental evidence, and select compounds in early drug discovery.
Apple is working on its own AI agent that can perform a range of tasks for you, including browsing the web, working across apps, writing code, and calling services. That's according to a now-removed job listing spotted by Bloomberg, which s
Traders work on the floor of the New York Stock Exchange during morning trading on November 18, 2026 in New York City. | Image: Getty Images "We've got this," Colette Kress, interim CEO of Nvidia, said Tuesday night as she closed out the
This week on “Uncanny Valley,” we discuss the outages that have swept digital life over the past week, where new money from Anthropic’s IPO is going, and the AI chip world post-Jensen Huang.
In Taiwan, some food vendors are distancing themselves from Nvidia’s former CEO. Others stand by him. “It’s not as if he killed someone or set something on fire,” says one shop owner.
Lawsuit: Over 99% of school system’s AI gun alerts were false positives.
The company’s alternate-universe internet serves as a massive sandbox that allows the Chrome team to properly train AI browsing agents—without the risk of them going rogue.
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. A human liver has been kept alive outside the body for a record seven weeks Researchers in Australia c
Researchers in Australia claim they’ve kept a human liver alive, on life support, outside the body for seven weeks, opening up speculative new possibilities in biomedical engineering. The feat, whose details are pending publication, was car
After a tough few years littered with some high-profile failures, there’s some positive news for next-generation batteries in the US. Sila just signed an agreement with Panasonic Energy, a global battery manufacturer, to supply silicon-carb
Learn how OpenAI built the Performance Factory, a system of Codex-powered workflows that finds and fixes performance issues to help keep ChatGPT fast.
Watch a demo of the Codex Security Plugin and see the workflow in action, from getting started to reviewing the results shown in the recording.
Watch a walkthrough of PR scans and see how the presenter reviews scan results in a pull request workflow.
Watch a walkthrough of onboarding to Daybreak Blue, with a closer look at the setup experience and how to get started.
Explore a reference architecture for supervised security investigations with Daybreak Red, including safety controls, setup, and monitoring.
Erratic Muse outputs sharing random bank, bitcoin balances aren’t real, Meta says.
With MyMems, Wright says he wants to "turn the simulation inward."
Like a nasty rash, AI seems to spread everywhere. Google crams AI Overviews down your throat even though it costs the company billions of dollars in lost ad revenue. Leave your apartment, and you might pass two people heaving an AI-powered
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. This smell startup says its AI nose is as good as a human’s In a New Jersey lab, researchers are tryin
In a lab near Newark Airport in New Jersey, a team of around 30 experienced human sniffers are slowly working their way through every odor on Earth. Their careful notes—“nutty,” “peachy,” “marine”—are cross-referenced with a chemical analys
Dean Ball, Head of Strategic Futures at OpenAI, examines how more capable AI could reshape society and how people can retain control of their institutions.
Sumitomo Life uses ChatGPT Enterprise to research insurers, prepare proposals, and draft code for investment analysis. Early use by executives helped drive adoption.
OpenAI measures how coding agents are accelerating AI research, where their performance remains uneven, and what this signals about recursive self-improvement.
See how Codex runs commands on your machine while enforcing the filesystem and network limits you control across macOS, Linux, and Windows.
Word, Excel, and PowerPoint now live inside Copilot; always-on agents are next.
“I love our country”: Jensen Huang’s apology comes amid escalating Nvidia probes.
Listen to the session or watch below In an effort to reach an elusive continuity of the self, some far-thinking technologists are working on ways to keep your brain going—even beyond death. Science editor Mary Beth Griggs and senior editor
The US Southwest generated 28% less electricity via hydropower between April and August than the 10-year average, according to new data from the US Energy Information Administration. Most hydropower plants in the US rely on dams to hold bac
AI training and chat history will now have separate toggles in the Gemini activity settings.
Ars Live discussion is happening today, November 17, at 1 pm Eastern Time.
AI agent takes unapproved actions even when laptop lid is closed and told not to.
Robotics startup misses payroll a year after CEO used funds for luxury home.
Predicting how DNA sequences and their variations impact specific human cell types and disease.
AlphaProof Code agents wrote three million lines of proofs in Lean to verify Signal’s core encryption.
Foundational precipitation and geostationary satellite datasets to observe the planet clearly.
Our most capable vision-language-action model enhances robot dexterity, reliability, and safety
Learn how Google Antigravity and our Cheminformatics Skills accelerate drug discovery workflows, search compounds, and propose novel rigidified macrocycles.
Google DeepMind and UK government work together to reduce planning assessment times