Skip to content

Scan record

Everything the site collected, newest first by collection time (Beijing time), back seven days.

Nov 1740 loaded

  1. SkillSpector TP1-TP4 findings are mislabeled as AML.T0080; tool poisoning should be AML.T0110

    NewsNVIDIA/SkillSpectorEventThe item you came from

  2. Google GTIG says it has observed fully automated exploit attacks for the first time, currently at Level 5

    NewsThe DecoderEvent

  3. Altimeter's Gerstner says Nvidia's board should keep the option of Jensen Huang's return open: "get this right"

    NewsSemaforAI updates

  4. OpenAI wants to stop deepfakes from overwhelming child abuse reporting hotlines

    NewsThe Verge AITools

  5. GPT-6 Astra helps produce a proof that may solve a 50-year problem in particle physics

    NewsThe DecoderAI updates

  6. Microsoft researchers propose hiding payloads in dormant neurons of open-weight models

    NewsDark ReadingAttacks

  7. Multiple banks restrict third-party AI agents from operating customer accounts

    NewsAnquan NeicanGovernance

  8. Report says attackers used an autonomous LLM framework; Google launches AI Threat Defense

    NewsGoogleEvent

  9. U.S. Department of Commerce plans to restrict China's remote access to AI chips

    NewsArs Technica AIGovernance

  10. U.S. Department of Commerce tightens controls on remote access to AI chips, closing a loophole for overseas data center access

    NewsNPRGovernance

  11. The Gemini 4 safety report fails to prove that the risk of loss of control is low

    NewsLessWrongOpinion

  12. The author says they are deeply uncertain about whether to urge friends at frontier AI labs to resign

    NewsLessWrongOpinion

  13. U.S. State Department official says the U.S. will not restrict technology companies headquartered in allied countries

    NewsSemaforGovernance

  14. The open-source movement wants Jensen Huang to return to Nvidia as CEO

    NewsMIT Technology ReviewGovernance

  15. Transluce says it will pilot a six-week residency evaluation with OpenAI starting in early December

    XX @TransluceAIGovernance

  16. Anthropic acknowledges the risks AI poses to humanity, but is its governance really up to the task?

    NewsThe Guardian · AIGovernance

  17. Nvidia's board is reportedly not considering asking Jensen Huang to return as CEO

    NewsThe Verge AIAI updates

  18. Altimeter's Brad Gerstner plans to invest $5 billion in Anthropic's IPO

    NewsCNBC · TechnologyAI updates

  19. OpenAI signs agreements with four independent evaluation organizations to strengthen safety work

    NewsCNBC · TechnologyGovernance

  20. UK Ofcom's new child safety rules do not include client-side scanning

    NewsThe RecordGovernance

  21. Bucks County, Pennsylvania students attend "deepfake" bill signing ceremony with the governor and Representative Marcell

    NewsPatchGovernance

  22. Trump says he "doesn't mind" Jensen Huang returning as Nvidia CEO

    NewsThe Irish TimesAI updates

  23. OpenAI CEO Sam Altman on Nvidia without Jensen Huang: the industry will be "fine"

    NewsTechCrunch AIAI updates

  24. Sources say Nvidia's board is not considering bringing Jensen Huang back as CEO

    NewsCNBC · TechnologyAI updates

  25. LangGraph advisory: MongoDBStore query expression injection, possible out-of-bounds data reads in shared multi-tenant scenarios

    Newslangchain-ai/langgraphjsEvent

  26. OpenAI: CSAM image classifier precision is lower in online evaluation than offline

    NewsOpenAI Deployment SafetyEvaluation

  27. "Heroes also have flaws": Sam Altman says he has personal-level empathy for former Nvidia CEO Jensen Huang

    NewsSemaforAI updates

  28. Redwood believes the Gemini 4 safety report does not prove the risk of loss of control is low

    NewsRedwood ResearchOpinion

  29. FBI says hackers used the open-source AI tool Artex to breach networks, with some cases compromised in under three hours

    NewsTechCrunch AIEvent

  30. U.S. to close AI chip loophole before Trump–Xi meeting

    NewsTIMEGovernance

  31. Midas Project investigation: over 80 X influencers may have been paid to endorse pro-AI articles, some involving Nvidia chips for China

    XX @themidasprojGovernance

  32. The Midas Project said it found a covert influence operation using right-wing creators to promote sales of Nvidia chips to China

    XX @themidasprojGovernance

  33. OpenAI releases a free tool for vetted organizations to detect child sexual abuse material, including AI-generated images

    XX @OpenAINewsroomTools

  34. Apollo Research says it has launched an embedded evaluation pilot with OpenAI; six evaluators will be stationed there for four months

    XX @MariusHobbhahnEvaluation

  35. METR and OpenAI reach an agreement to conduct a limited alignment evaluation pilot, focusing on OpenAI's largest-scale RL run

    XX @METR_EvalsEvaluation

  36. Google DeepMind says researchers use Gemini 4 Carbon in Antigravity to handle multi-step problems

    XX @GoogleDeepMindAI updates

  37. Apollo Research and OpenAI launch an embedded evaluation pilot to investigate whether models attempt to undermine oversight of themselves

    XX @ApolloResearchEvaluation

  38. Apollo Research and OpenAI launch an embedded evaluation pilot, with a six-person team stationed for four months to evaluate model alignment

    NewsApollo ResearchGovernance

  39. GTIG AI Threat Tracking Report: Agentic threat activity crosses a critical threshold

    NewsGoogle Cloud BlogEvent

  40. Unit 42 discloses that China-linked attackers breached 10 organizations with AI-driven attacks

    NewsDark ReadingEvent