Skip to content

Scan record

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

Nov 1740 loaded

  1. Beyond Patches: Control-Gated Benign Probe Reconstruction of Patch Bypasses from Incomplete Fixes in LLM Agent Frameworks

    PaperarXiv: 其他 AI 安全论文The item you came from

  2. AudioTrace: File-Level Membership Detection for Audio LLMs

    PaperarXiv: 其他 AI 安全论文

  3. Swarm Shepherd: Protecting Multi-Agent Ecosystems from Persistent Latent Intrusions

    PaperarXiv: 其他 AI 安全论文

  4. German research reveals that at least 7,403 databases of apps developed by users on the AI development platform Lovable have improper permission settings

    NewsiThome (Taiwan)Event

  5. InclusionMed: An AI benchmark for inclusive healthcare

    PaperHugging Face Daily Papers

  6. Belo Horizonte City Council to hold hearing on November 18 to discuss AI in public safety

    NewsPortal CMBHGovernance

  7. DominoBench: evaluating multi-step physical reasoning from static images

    PaperHugging Face Daily Papers

  8. SWE-CoWork: Benchmarking Coding Agents as Digital Employees in Organizations

    PaperHugging Face Daily Papers

  9. Centered Ablation: why last-layer direction ablation is insufficient or collapses, and a closed-form correction

    PaperHugging Face Daily Papers

  10. A quick guide to data center construction ban litigation

    NewsFinancial Times · AIAI updates

  11. OpenAI uses AI to strengthen cyber defense for NATO and allied countries

    NewsOpenAIEvent

  12. According to sources, Nvidia warned chip buyers that remote access to AI chips from China may bring legal risks

    NewsCNBC · TechnologyGovernance

  13. Apollo Research member: training against CoT-like surfaces may be similar to training against CoT

    NewsLessWrongAlignment

  14. Australia and 10 other countries signed the US-led San Francisco Declaration to protect the sensitive technologies and supply chains needed for advanced AI

    NewsACSC (Australia)AI updates

  15. Helen Clark: Automated warfare has arrived, and governments must restrain it

    NewsThe Guardian · AIGovernance

  16. Hackers use AI to automate vulnerability research, and exploited zero-day vulnerabilities surge to 22 in a single month

    NewsiThome (Taiwan)Event

  17. OpenAI opens access to 4 external organizations to examine AI model safety in depth

    NewsiThome (Taiwan)Governance

  18. The U.S. Department of Commerce and the Department of Justice said that in 2024 they had opened an investigation into Nvidia partner Megaspeed

    NewsCNBC · TechnologyGovernance

  19. Google report: threat actors use AI to launch cyberattacks, often without human involvement

    NewsHindustan TimesEvent

  20. AI security accounts for more than one-third of cybersecurity seed-round financing deals

    NewsHelp Net Security AIAI updates

  21. K2-Horizon: an open-source language model frontier fleet

    PaperHugging Face Daily Papers

  22. Malware found in the @shadowdqj/image-mcp package

    NewsGitHub Security AdvisoriesEvent

  23. Palo Alto Networks report: 40% of organizations struggle to trace AI agents' actions

    NewsHelp Net Security AIDefenses

  24. GitHub gh-aw weekly: v1.2.3 pre-release, default model routing favors cheaper models on Copilot

    NewsGitHub Agentic WorkflowsAI updates

  25. Malware found in the @shadowdu/memos-mcp npm package

    NewsGitHub Security AdvisoriesEvent

  26. The U.S. Department of Commerce Bureau of Industry and Security issued an interim final rule to prevent remote access to U.S. superintelligence chips via third-country data centers from circumventing export controls.

    Newsbis.govGovernance

  27. US Commerce Department and Justice Department issue joint statement on Megaspeed investigation

    XX @MinLiBuildsGovernance

  28. GHSA flags mem0-mcp-server as a malicious package

    NewsGitHub Security AdvisoriesEvent

  29. US Senator Cantwell sent a letter to Commerce Secretary Lutnick questioning the enforcement of AI chip export controls

    NewsU.S. Senate Committee on Commerce, Science, and TransportationGovernance

  30. The AI doomer author says: if ASI has not appeared by 2028, his camp will be respected even less

    NewsLessWrongOpinion

  31. The $900bn data center bet behind Anthropic's listing

    NewsFinancial Times · AIAI updates

  32. Lex column: The value of Chinese robot makers goes beyond the backflip videos flooding social media

    NewsFinancial Times · AIAI updates

  33. As public backlash intensifies, banks seek to reduce AI debt exposure

    NewsFinancial Times · AIAI updates

  34. FT View: Do we really want work to become machine-readable? Concentrating enterprise information into AI systems may redistributively reallocate power

    NewsFinancial Times · AIOpinion

  35. Shadowserver observation: 3,847 instances of AI attack frameworks exposed on the public internet, 20 in Taiwan

    NewsiThome (Taiwan)Event

  36. United States expected to say no to a global ban on killer robots

    NewsNewsweekGovernance

  37. Why does a scam only reveal itself after a few messages? Singapore GovTech AntiScam Pango uses AI to analyze the entire conversation and identify scam patterns

    NewsGovernment Technology Agency of SingaporeTools

  38. 16 governments and 10 AI companies sign the Paris Call for AI Safety

    Newsuk.diplomatie.gouv.frGovernance

  39. Federal Reserve Governor Barr discusses several potential effects of artificial intelligence on financial stability

    Newsfederalreserve.govGovernance

  40. Vulnerabilities in multiple IBM products; Canadian Centre for Cyber Security issues security advisory AV26-1162

    NewsCanadian Centre for Cyber SecurityEvent