List viewFlip view
AI at the Threshold
Weekly Release

AI at the Threshold

16 November 2026
  1. 1
    Industry

    Apple launches study of clinically grounded health AI

    Apple and OpenEvidence have launched a U.S.

    Apple launches study of clinically grounded health AI
  2. 2
    Research

    Physicist and AI model propose lattice definition of the Standard Model

    Physicist Nathaniel Craig says he used OpenAI’s GPT-6 Astra to develop a candidate nonperturbative lattice definition of the Standard Model.

    Physicist and AI model propose lattice definition of the Standard Model
  3. 3
    Business

    Anthropic IPO Puts AI Market Sentiment to the Test

    Anthropic’s planned stock-market debut at a reported $1.4 trillion valuation comes amid market turbulence and political opposition affecting the AI industry.

    Anthropic IPO Puts AI Market Sentiment to the Test
  4. 4
    Tools

    New Datasette Agent release adds question input to the homepage

    Version 0.6a0 of datasette-agent has been released with several minor fixes.

    New Datasette Agent release adds question input to the homepage
  5. 5
    Regulation

    US lawmakers move to tighten AI chip export controls on China

    US lawmakers are accelerating bipartisan proposals to restrict China’s access to advanced artificial intelligence chips following revelations involving Nvidia founder Jensen Huang and a Singapore-based customer.

    US lawmakers move to tighten AI chip export controls on China
  6. 6
    Safety

    South Korean tax exam candidate admits using AI to cheat

    A candidate who received the highest score on South Korea’s national tax adviser examination admitted using generative AI to answer questions.

    South Korean tax exam candidate admits using AI to cheat
  7. 7
    Model launches

    Google introduces Gemini 4 Hydrogen as lower-cost execution model, but pricing sits close to flagship siblings

    Google today unveiled Gemini 4 Hydrogen, the lowest-cost member of the Gemini 4 family and the new free-tier model that will power AI features in Google Search.

    Google introduces Gemini 4 Hydrogen as lower-cost execution model, but pricing sits close to flagship siblings
  8. 8
    Industry

    AI could reshape tasks for most Hungarian workers, labour market map finds

    A joint BeHive and AI Budapest project estimates that AI could significantly change the work of about 64% of Hungary’s employees over the next two to three years.

    AI could reshape tasks for most Hungarian workers, labour market map finds
  9. 9
    Research

    Opinion predicts growing prominence for TerminalBench task authors

    According to one opinion, because TerminalBench v4.1 has only 66 tasks and has attracted interest from leading artificial intelligence laboratories, the authors of its test tasks may eventually receive more recognition than researchers publishing in Nature or Science.

    Opinion predicts growing prominence for TerminalBench task authors
  10. 10
    Business

    Bharat Semi raises $24 million to develop AI image sensors

    Indian startup Bharat Semi has raised $24 million in seed funding to develop programmable image sensors that combine image capture, processing and AI.

    Bharat Semi raises $24 million to develop AI image sensors
  11. 11
    Tools

    OpenAI launches science blog on research uses of AI

    OpenAI is launching a science blog for researchers to share findings, methods and experiences involving artificial intelligence.

    OpenAI launches science blog on research uses of AI
  12. 12
    Regulation

    EU AI content marking grace period approaches December deadline

    A grace period for certain AI systems already available in the European Union is due to end on December 2, 2026, according to EY.

    EU AI content marking grace period approaches December deadline
  13. 13
    Safety

    Anthropic launches Claude Opus 4 under ASL‑3 security and deployment standards

    On May 22, 2025 Anthropic announced that Claude Opus 4 is being deployed with AI Safety Level 3 (ASL‑3) Deployment and Security Standards from its Responsible Scaling Policy.

    Anthropic launches Claude Opus 4 under ASL‑3 security and deployment standards
  14. 14
    Model launches

    Gemini 4 Carbon release: many advances, but reliability issues

    Gemini 4 Carbon has been released; the model shows significant improvements but is slow, clumsy and often gives incorrect answers (hallucinates).

    Gemini 4 Carbon release: many advances, but reliability issues
  15. 15
    Industry

    European firms trail US peers in turning AI adoption into benefits

    A 2026 European Investment Bank survey finds that EU businesses are close to US firms in generative AI adoption but lag in realizing its benefits and investing in existing technologies.

    European firms trail US peers in turning AI adoption into benefits