Today's trail

China's access to some Nvidia chips in doubt as curbs gain traction in US Congress

By David Morgan, Courtney Rozen and Mike Stone Nov 13 (Reuters) - Chinese AI shares tumbled on Friday, as investors worried that Jensen Huang's departure as Nvidia CEO could further restrict Chinese access to advanced US chips,

Open Source AI Models: 3 Run Paths, 48% Coding Index [2026]

Open source AI models compared: Kimi K3.1 Pro, DeepSeek V4.1 Flash, Mistral Large 4, GLM-5.3, Gemma 4. License evidence, prices from $0.15, 3 run paths.

Jensen Huang Exit Sets Up US-China Showdown Over AI Chip Curbs

The fall of Nvidia Corp. Chief Executive Officer Jensen Huang has removed the industry's most prominent opponent of US chip controls, setting up a showdown over potential new curbs that could accelerate China's drive for semiconductor self-sufficiency.

Post-Jensen US Chip Curbs Threaten Southeast Asia's AI Boom

Southeast Asia risks facing tighter US regulations on the export and use of high-powered chips, threatening to stall the region's AI-fueled economic boom.

Nvidia's rivals, from AMD to Google, and the chips they are betting on to loosen its AI grip

Nvidia supplies about three-fifths of AI computing power. How AMD, Google, Amazon and Huawei compete, and why gains may come slowly.

Google launches Gemini 4 Hydrogen AI model: Know features, price, and more

United Airlines bans AI agents from booking flights

Anthropic’s 2 Frontier Models Follow Slowdown Call [2026]

Google AI glasses start at $395 and ship 10 December. What Gemini does on them, which iPhones work, what iPhone owners miss, and how they compare to Meta.

TO THE NEW Partners With SpaceXSI To Bring AI Coding Platform Cursor Into Enterprise AI Workflows

Singapore-headquartered digital engineering firm, TO THE NEW, has been named a SpaceX Super Intelligence Trans

Lightweight models take center stage, and the conditions for running agents are questioned | 6 AI news stories for November 13

Hello, this is Yui / Yuichi.Looking at today's news, the focus is less on the performance race of large models and more on how to choose light, fast models, and under what conditions to run AI agents.