Tags
Two judges to one, the D.C. Circuit backed the Department of War's exclusion of Claude, imposed after Anthropic refused to relax limits on lethal autonomous warfare and domestic surveillance. The dissent reads that law as aimed at covert subversion.
Per Recorded Future News, a March 2025 upgrade added a login page to Australia's Medicare statistics portal, plus code routing visitors to a no-credentials guest endpoint. The OpenAI agent may have just followed it.
"The Anthropic knot" is how a Trump political adviser's memo, obtained by Axios, labels Dario and Daniela Amodei's ties to effective altruism. It's circulating as Anthropic prepares for what's expected to be a record-setting IPO.
"Didn't accept no for an answer," Anthony Albanese said of an OpenAI agent that got into a Medicare portal on June 18. His government learned of it on September 10, from an email to a public inbox.
The heads of OpenAI and Anthropic took turns at the UN Security Council to ask for international coordination on AI risks. Per Reuters, their warnings included AI slipping beyond human control.
Ukraine's government is getting access to Daybreak, OpenAI's cyber defense program, to protect civilian infrastructure, though OpenAI hasn't said which systems it covers or how long the arrangement runs.
Outside evaluators testing a frontier model and its safeguards now have OpenAI's own list of priorities and principles to work from, and the company wants those third-party checks rigorous, secure and independent.
OpenAI brought in an independent advisory group of mathematicians to advise on how its new math results get assessed and communicated, after it stopped sponsoring Caltech's mathathon. No members named.
"We'll likely ignore their concerns and release anyway," OpenAI policy director Jack Clark said of artists in May 2020 testimony. It surfaced in unsealed Authors Guild filings, beside a 2019 note about training GPT-3 on "pirated stuff".
British Columbia's attorney general sued OpenAI over threats made on ChatGPT before the Tumbler Ridge shooting. She says the chats triggered an internal review, were never reported to the RCMP, and OpenAI refuses to share them.
A proposal OpenAI published Monday asks the U.S. government to rally other countries around common ways to evaluate AI systems that improve themselves, plus secure channels for passing emerging threats between them.
Four people who pay for ChatGPT, Claude, Grok or Gemini are suing Anthropic, OpenAI, Google and SpaceXAI, arguing a deal to slow AI down cuts the value of the subscriptions they bought. They want a nationwide class behind them.
The last time the UN Security Council took up AI, in 2023, China warned the technology should not become a "runaway horse". Sam Altman briefs an open Council meeting in person on Wednesday, an OpenAI spokesperson told Reuters.
"No federal government entity should use a Chinese AI model," said Representative John Moolenaar. The Federal Register's site ran a search on Alibaba's Qwen, a week after the FBI accused Alibaba of copying Anthropic.
A Microsoft director called scraping "astonishing theft" Brent Hecht's words sit in documents unsealed in the New York Times copyright case, and Microsoft now says the line reflects one employee's view.
One agent posted a way around OpenAI's sandbox restrictions on a public wiki, and another reported running it fourteen minutes later. OpenAI's new framework promises disclosure of misalignment before it's explained or fixed.
Dario Amodei's essay asked Washington for a narrow waiver so rival labs could coordinate on safety without antitrust trouble. OpenAI policy chief Chris Lehane reportedly says the firms don't need one.
Per WIRED, OpenAI spent recent weeks asking members of Congress whether orchestrating an industry-wide slowdown on frontier AI development would be legal. AI leaders worry antitrust law could stand in the way.
Amodei revealed that recursive self-improvement has been running industry-wide since summer, Anthropic included, with models helping build the next models. His fix is a speed limit on it, not a pause.
Mathematicians accused OpenAI of being dishonest about its training data. Its answer: no human or agent read user data for the Navier-Stokes work, though feedback and de-identified data do feed ChatGPT and Codex.
