ai
170 stories filed under this topic.
-
Claude Beat the Best Published Quantum Circuits by Up to 17x
Anthropic gave Claude Mythos 5 plain-language specs for eight quantum circuits and an automated scorer. Working largely on its own, it beat the best published designs on all eight — including a decoder 12x cheaper than Google's.
-
An OpenAI Agent Asked Itself Whether Taking a Stranger's Crypto Was Ethical — Then Did It Anyway
OpenAI disclosed that a GPT-5.6 Sol research model used a public recovery phrase to spend from a stranger's wallet on Ethereum's Sepolia testnet — after asking itself whether that was ethical, and then leaving the whole episode out of its report.
-
Wall Street Wants AI Agents Out of Your Bank Account — Unless the Bank Built Them
JPMorgan, Charles Schwab, and Citi all moved this week to block outside AI agents from customer accounts — while building their own agents behind the same walls.
-
Amazon's Bee Bracelet Listens to Your Whole Day — and Charges $11.99 a Month to Remember It
Amazon's new Bee Frontier bracelet and Bee Eclipse phone clip listen all day and turn conversations into summaries and to-dos. The hardware starts at $75 — the long memory is a subscription.
-
Apple Rewrites Its App Review Rules to Finally Mean It on Nudify Apps
Apple updated its App Review Guidelines on Friday with sharper penalties for AI "nudify" apps and bait-and-switch review evasion — including developer account termination and blocking offenders from alternative distribution.
-
ChatGPT's Chrome Extension Can Now Read and Control Your Open Tabs
OpenAI's updated Chrome extension lets ChatGPT pull from the tabs you already have open, and lets ChatGPT Work fill out forms, send messages, and organize tabs for you.
-
China Proposes Export Curbs on 23 AI Technologies, Including Its Largest Models
Beijing wants to add 40 AI, biotech, and solar technologies to its export-control list, and exporters have to register transfers starting today.
-
Cody Wilson Test-Fires What He Calls the First AI-Made Gun
Defense Distributed says its custom model, GatGPT, designed a 3D-printed .22 pistol called the WASP, and The Verge watched it fire two of three shots.
-
India Says AI Agents From Multiple Labs Read Personal Data at Local Companies
India's cybersecurity agency says AI agents under testing by several model providers got into live systems at Indian organizations and read personal data and confidential records.
-
Meta's New Business Terms May Let It Train AI on News Sites for Free
Updated commercial terms taking effect October 30 authorize Meta to crawl websites associated with business accounts and use that content to train AI products.
-
Replit CEO Says It's No Longer Working With Anthropic
Amjad Masad said onstage at TechCrunch Disrupt that Replit has moved on from Anthropic over cost and performance, though a person familiar with the deal says the partnership is still in place.
-
Anthropic Gives METR Employee-Level Access to Audit Its AI Safety
Six METR researchers now sit inside Anthropic with hardware and tool access on par with the company’s own risk teams, and Anthropic cannot veto what they publish.
-
Apple Says It Found No Trace of OpenAI's Files on Its Returning Engineer
Apple's new court filing says OpenAI raised its file-theft claim without notice, that Apple's search turned up nothing, and that OpenAI went to the judge instead of answering Apple's follow-up questions.
-
Greg Joswiak Says Apple Doesn't Feel Threatened by OpenAI's Hardware Plans
Apple's marketing chief told TIME the company isn't building AI for the "lunatic fringe" chasing superintelligence, and that its integration gives it the edge over OpenAI devices.
-
Meta Releases Muse Spark Open Weights, Tuned to Run on Two Mac Studios
Meta calls the 816B-parameter model the most capable US open-weight model yet, and it worked with Apple to optimize it for MLX on M5 Ultra.
-
Nvidia Releases Nemotron 3.5 Super VL and Bolt for Open Agents
The chipmaker’s new open agent models span a 120B multimodal Super VL for servers and a 74B Bolt that fits in 24GB for local RTX and Spark systems.
-
OpenAI Offers U.S. University Labs Free GPT-Rosalind Access for a Year
Eligible U.S. academic labs can get up to 30 free Business Standard seats for 12 months, with GPT-Rosalind after trusted-access approval; API access is not included.
-
Ramp Raises $1.85B at a $60B Valuation, Four Months After Hitting $44B
The expense management company says annualized revenue has passed $2 billion as it piles AI agents into procurement, accounting and payments.
-
Volvo Expands Its Nvidia Deal for Hands-Free Urban Driving
Volvo will use Nvidia’s Hyperion and Alpamayo stack for Level 2++ point-to-point driving on city streets, aiming to launch before 2030.
-
40 Attorneys General Want Protections on Character.AI's Chat Logs Before Any Sale
A day after Character.AI filed for Chapter 11 with Disney bidding $20 million for a license to its tech, a bipartisan group of 40 attorneys general told the court any buyer has to protect kids and users' private chats.