Skip to content

Scan record

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

Nov 1839 loaded

  1. OpenAI's "independent evaluation" does not look independent at all

    NewsMother JonesGovernanceThe item you came from

  2. Microsoft's agent-framework PyPI package was compromised and implanted with credential-stealing code

    NewsSocketEvent

  3. Anthropic reportedly plans to raise its IPO issue price

    NewsSemaforAI updates

  4. Nvidia says it does not plan to reappoint Jensen Huang as CEO

    NewsTechCrunch AIAI updates

  5. Nvidia says its open-source AI strategy remains unchanged after Jensen Huang's resignation

    NewsTechCrunch AIAI updates

  6. Wiz: malicious versions 1.26.1/1.26.2 of Microsoft agent-framework steal Azure and CI credentials

    NewsWiz ResearchEvent

  7. Nvidia's board has no intention of letting Jensen Huang return as CEO

    NewsWIRED Security and AIAI updates

  8. Apple launches new security controls for Mac to address AI agent risks

    NewsFredzoneDefenses

  9. Rogo evaluation: GPT-6.2 Astra financial accuracy 61%, surpassing Opus 5.5

    NewsRogoAI updates

  10. California Attorney General investigates OpenAI over the Hugging Face attack

    NewsBBC NewsGovernance

  11. Calif Research says it used AI to find the OpenJDK vulnerability Java4Shell

    NewsCyberScoopEvent

  12. UN's top tech envoy says US-China AI race is triggering global "alarm"

    NewsSemaforGovernance

  13. Nvidia's quarterly revenue doubles; board chairman says no intention to reinstate Jensen Huang as CEO

    NewsThe Guardian · AIAI updates

  14. The Wall Street Journal reports that the SEC is investigating whether Nvidia's disclosures misled investors.

    NewsTechCrunch AIGovernance

  15. Nvidia without Jensen Huang: Can the chip giant survive the CEO resignation crisis?

    NewsFinancial Times · AIAI updates

  16. German official says Europe needs "innovation-friendly" regulation to catch up in the AI race

    NewsSemaforAI updates

  17. Redwood Research proposes capability activity definition standards for limiting frontier AI progress by compute

    NewsRedwood ResearchGovernance

  18. How Google joined the tech outage wave: by announcing that Gmail had no outage

    NewsTechCrunch AIAI updates

  19. Anthropic recommends upgrading to Claude Fable 5.2 only for the hardest technical projects

    NewsLessWrongAI updates

  20. Why do AIFP forecasters diverge so much on the SAR and TED-AI gap?

    NewsLessWrongOpinion

  21. Amid the Nvidia controversy, the US tightens controls on remote access to AI technology

    NewsWPRLGovernance

  22. Varonis: TheMailer phishing kit uses LLMs to automate phishing end to end, rewriting lures until they bypass spam filters

    NewsDark ReadingEvent

  23. European Commission opens public consultation on the draft template for fundamental rights impact assessments under the Artificial Intelligence Act

    NewsEuropean CommissionGovernance

  24. GPT-6.2 Astra scores 174 on the Epoch Capability Index, tied with Fable 5.2

    XX @EpochAIResearchAI updates

  25. OpenAI auditor clarifies: only company-approved employees, no quoting or summarizing

    XX @BethMayBarnesGovernance

  26. Even Sam Altman can't escape AI deepfakes on Instagram

    NewsThe Verge AIAI updates

  27. Billboard accuses Trump and Jensen Huang of collusion; critics oppose Jensen Huang's return to Nvidia

    NewsArs Technica AIGovernance

  28. OpenAI releases GPT-6.2 Astra and Sol, with lower API prices and improved scores on some benchmarks

    NewsThe DecoderAI updates

  29. How to securely expose Amazon Bedrock AgentCore Runtime externally

    NewsAWS SecurityTools

  30. Widespread outage hits banks, air travel, NSA and Apple, disrupting almost everything

    NewsCNBC · TechnologyEvent

  31. PyPI package mcp-aevo-server contains malicious code, steals credentials from environment variables and listens for remote commands

    NewsGitHub Security AdvisoriesEvent

  32. GPT 6.2 Astra system card shows chain-of-thought-free problem-solving capability continuing to rise; researchers say it likely reduces chain-of-thought monitorability

    XX @tomekkorbakEvaluation

  33. Irregular evaluates GPT-6.2 Astra with offensive security benchmarks

    NewsIrregularEvaluation

  34. U.S. issues new rule after Nvidia controversy restricting China's remote use of advanced AI chips via third-country data centers; China criticizes the move

    NewsBusinessLineGovernance

  35. CISA Acting Director Nick Andersen: whether or not a cyberattack is involved, a string of technical outages highlights the need for system resilience

    NewsCyberScoopGovernance

  36. Researchers say they demonstrated bypassing Telegram Android's new filter on a test server

    NewsThe Hacker NewsEvent

  37. Spanish Data Protection Agency President argues GDPR Article 22 automated decision-making should move toward a clearer legal framework, prioritizing effective safeguards over general prohibitions

    NewsIlustre Colegio de la Abogacía de MadridGovernance

  38. Fei-Fei Li says universities and the public sector are locked out of AI because it is too expensive

    NewsSemaforOpinion

  39. Microsoft Perception officially released: red, blue and green team security agents in closed-loop collaboration

    NewsMicrosoft SecurityTools