// AI Briefing · November 12, 2026

AI Briefing

Today’s video is on its way

The stories are ready to read now. The video usually follows within the hour.

Rendering the short

Nvidia's board met late Wednesday to decide whether to remove CEO Jensen Huang, after the Wall Street Journal reported it was preparing to oust him and the New York Times reported he had an affair with an executive who reports to him. Anthropic launched Claude Sheets and Claude Whiteboards in beta for paid subscribers and made its Dashboards data tool generally available, taking a direct run at Microsoft's Office franchise. New York opened the registration portal for its RAISE Act, which requires frontier AI developers with more than $500 million in annual revenue, such as OpenAI and Anthropic, to register by January 1, 2027. Google DeepMind announced Gemini Robotics 2.5, a vision-language-action model that it says beats previous baselines by 30% to 100% depending on the task. Security firm Claroc found that 35,578 of 37,571 live apps built with SpaceXSI's Grok Build could be opened without signing in. UK chip startup Fractile raised a $720 million Series C from Thrive Capital, Founders Fund, Lightspeed, Redpoint and Britain's Sovereign AI Fund.

Nvidia's Board Meets on Whether to Oust Jensen Huang After a Second Affair Report

Nvidia's board met late Wednesday to decide whether to remove CEO Jensen Huang, after the Wall Street Journal reported it was preparing to oust him and the New York Times reported he had an affair with an executive who reports to him. This follows his admission on Sunday of a relationship with a woman US intelligence reportedly believes is a Chinese spy, and no decision had been announced by Thursday morning in Europe. Nvidia shares fell about 8% after hours on Wednesday. House Republicans have also asked Huang for a closed-door briefing, so the world's most important AI chipmaker faces a leadership vacuum with no named successor.

02ProductYahoo Finance (GuruFocus) ↗

Anthropic Pushes Into Office Software With Claude Sheets and Whiteboards, and Microsoft Shares Fall

Anthropic launched Claude Sheets and Claude Whiteboards in beta for paid subscribers and made its Dashboards data tool generally available, taking a direct run at Microsoft's Office franchise. Sheets drafts and revises spreadsheets, marks each cell it changes, and exports to .xlsx or CSV, though Anthropic says the functionality is basic for now. Microsoft shares fell about 2% on Tuesday and slid again Wednesday as investors weighed the threat to Microsoft 365. Microsoft is also an investor in Anthropic, and the launch comes days before Anthropic's IPO.

03PolicyTechCrunch ↗

New York Opens Registration Under Its RAISE Act, Bringing State Oversight to Frontier AI Labs

New York opened the registration portal for its RAISE Act, which requires frontier AI developers with more than $500 million in annual revenue, such as OpenAI and Anthropic, to register by January 1, 2027. Covered companies must publish safety frameworks and model transparency reports and report critical safety incidents within 72 hours. Fines reach $1 million for a first violation and $3 million for later ones. A new state office, DIGIT, will enforce the law, backed by a new Binghamton University center that will independently evaluate advanced AI systems. It is one of the most concrete state-level AI safety regimes in the US.

04ResearchGoogle DeepMind ↗

Gemini Robotics 2.5 Runs Humanoids on Mercedes-Benz Production Lines With a 96% Success Rate

Google DeepMind announced Gemini Robotics 2.5, a vision-language-action model that it says beats previous baselines by 30% to 100% depending on the task. In a proof-of-concept on Mercedes-Benz production lines, Apptronik's Apollo 2 humanoid sorted and packed car parts with a 96% success rate, by DeepMind's account. The trial covered one narrow kitting task, but it is a notable step for general-purpose robot models in real factories. The model is rolling out to robotics partners rather than as a public release.

05IndustryCloud Security Alliance ↗

Researchers Find About 35,500 Grok Build Apps Open to Anyone, Some Holding Medical and Corporate Data

Security firm Claroc found that 35,578 of 37,571 live apps built with SpaceXSI's Grok Build could be opened without signing in. These included a hospital staff roster, an ICU briefing with a patient's medical details, and 24 apps built by SpaceXSI and X staff, among them a confidential ad proposal to a US federal agency. Claroc estimates 650 to 800 of the open apps hold high- or critical-sensitivity data. SpaceXSI took down its own staff apps within about two days but had not addressed users' exposed apps. The finding shows the security risks of 'vibe-coded' apps that are public by default.

06FundingTechCrunch ↗

Britain's Fractile Raises $720M to Build Inference Chips With DRAM Stacked on Top

UK chip startup Fractile raised a $720 million Series C from Thrive Capital, Founders Fund, Lightspeed, Redpoint and Britain's Sovereign AI Fund. The UK government valued the company at over £5 billion, while the FT reported about $7 billion. Fractile stacks DRAM directly on its chips to cut the cost of AI inference and plans to ship its first chips and systems next year. The round makes it one of the best-funded challengers to Nvidia in inference hardware and is a sizable UK government bet on domestic AI hardware.