Hydrogen and ChatGPT Wallet expand AI-agent workflows, llama.cpp reaches 1.0, and new research highlights the difference between model behavior and enforced guarantees. NVIDIA also begins a leadership transition. This edition covers announcements before 17:00 UTC on 13 November.
AI image and video models, coding-agent infrastructure and security workflows gained more specific controls this week. The practical question is what each interface makes possible—and which conditions still need checking.
For teams using AI coding agents, Conductor's revised continuation statement and OpenAI's proposed Pierre acquisition show why ownership, ongoing service and a usable exit path need separate checks. Updated 13 November 2026 to reflect Conductor's changed notice.
NVIDIA introduced Nemotron for Science on 16 November, combining its existing Nemotron 3.5 Super VL AI model with scientific evaluations, a chemistry dataset and post-training recipes. The package is aimed at researchers building agents that interpret scientific information, generate code and call computational tools—not a newly released base-model …
Ouster announced on 17 November that Nashville is adding 100 intersections to its BlueCity traffic-management contract. The system pairs stationary 3D lidar with AI that tracks vehicles, cyclists and pedestrians for signal control. Ouster puts the total above 200 contracted locations; that is not a verified count of …
OpenAI says it is providing NATO institutions with advanced AI cybersecurity models through the alliance's Communications and Information Agency. Its November 17 update describes defender access and AI-agent training, not a new model release or an independently measured security outcome.
The Federal Reserve is examining how advances in AI could change productivity, jobs and financial risk. Kevin Warsh's November 16 prepared remarks present possible outcomes—not a forecast of a market crash—while a separate stability report examines financing behind advanced-computing infrastructure.
NVIDIA announced a four-system DGX Spark HPC Developer Kit at SC26 on November 16 for local AI models, agents and scientific applications. The company describes the kit as available for sale.
Google announced 66 additional regional voices for its Gemini AI assistant on Android and iOS on November 16. The selection applies across Live conversations and spoken chat responses.
OpenAI is expanding free access to its ChatGPT AI assistant for academic research. Announced November 16, the next round opens November 23 with 50,000 grants—not 200,000 new awards.
Lambda announced a co-packaged-optics network connecting more than 10,000 NVIDIA Blackwell Ultra GPUs at its Quebec campus on 16 November. The AI cloud operator says its tests found lower network latency and cooler optics than conventional pluggable designs, without increasing measured throughput.
NVIDIA reported Vera Rubin NVL72 results for AI training and agentic inference on 16 November. The company claims up to twice the training performance per GPU versus GB300 NVL72, but its highlighted inference-efficiency comparison uses a provisional Vera Rubin result that still awaits MLCommons verification.
Google is rolling out new planning tools for Ask Maps, its Gemini-based Maps assistant. The 16 November update adds itineraries and saved-list search where Ask Maps is offered, while route tips and appointment-finding start in the U.S. Availability differs by feature as the wider service expands across Europe.
NVIDIA announced Rubin HPC at SC26 on 16 November, describing a GPU designed to speed up the simulations and data processing that scientific AI agents use to test candidate ideas. The company expects availability in 2028, rather than presenting it as a product researchers can deploy today.
OpenAI launched a science blog on 15 November for researchers to explain their findings and AI-assisted methods. Its first guest post, by physicist Nathaniel Craig, describes GPT-6 Astra's contribution to an October proposal for defining the Standard Model on a spacetime lattice. Craig presents it as a candidate …
A November working paper from the Innovation for Cool Earth Forum (ICEF), covered by Axios on 15 November, calls for caution before deploying large constellations of orbital data centers to train and run AI models. It identifies potential harm to night skies, astronomy and the atmosphere, alongside collision …
JioStar says it used the OpenAI API to build conversational video and cricket discovery in JioHotstar. In an 11 November interview, its chief architect describes domain-expert evaluations and a separate model-based validator, while stressing that production monitoring needs more than fast infrastructure.
GitHub's 12 November availability report says an invalid credential at an upstream AI-model provider caused errors for several GPT-family models in Copilot on 28 October. Other models remained available, while features tied to one model were hit harder. This is a retrospective account of a resolved incident, not …
Holiday Extras says Codex, OpenAI's AI coding agent, helped its team reach a working mobile-app rebuild in two weeks. A 12 November OpenAI customer story reports better booking conversion and fewer errors afterward, but compares different periods rather than establishing that Codex alone caused the changes.
Amplitude describes Wave, a beta AI product agent that turns analytics, session replays, feedback and code into proposed improvements and implementation plans for coding agents such as Codex. Its 12 November OpenAI customer account reports a one-day homepage experiment and task-specific model trade-offs, rather than independently validated general …
OpenAI has pledged $2 million to a research network studying AI’s risks and opportunities for children. Its 12 November announcement supports shared measurement tools under the Coalition for Children’s Rights and Protection in the Age of AI; the linked charter sets research-independence and funding safeguards, rather than reporting …
Shield says GPT-6 Astra computer-use agents in its Forge platform now handle supported tasks in legacy IT portals, including systems without APIs. In a 13 November OpenAI customer account, the company describes live browser workflows while technicians decide which actions require review or approval; remote-desktop automation is still …
Google Research has detailed how Antigravity’s /teamwork combines AI agents with external checks for research and engineering. Its 13 November update reports new applications, while distinguishing tools available now from specialized workflows still to come.
Google is rolling out an optional Google Wallet connection for Gemini app users in the United States. The AI assistant can use spending information to answer budgeting questions, but Google says it cannot move, manage or invest money through the connection.
Ai2 has released XYBench, an evaluation of whether language-model assistants help with a user's underlying goal instead of reinforcing a mistaken approach. Its reported experiments distinguish mentioning a useful solution from making that solution the focus.
OpenAI introduced ChatGPT Wallet on 13 November for agents to complete purchases with user approval. Its initial rollout is in dots for ChatGPT Pro subscribers aged 18 or older in the US with eligible US-issued cards—not a general launch across every ChatGPT product.
The ggml-org project announced 1.0 releases for llama.cpp's libllama and the ggml tensor library on 13 November. The milestone gives developers embedding local AI inference explicit API and ABI versioning rules; it is not a new AI model or a universal performance claim.
GitHub's 12 November practical guide explains how AI coding-agent skills can share reusable instructions while tools, project configuration and permissions remain specific to each host. Its Data Forge and Proof Agent examples distinguish portable files from a working, enforced workflow.
IBM's Granite team released experimental LLMON previews on 12 November. The AI models use reserved structural tokens and explicit instruction-to-data bindings to distinguish tasks from supplied content—not an enforced security boundary.
Google introduced Gemini 4 Hydrogen on 12 November, an AI model aimed at everyday coding and multi-step agent work. The launch spans developer tools, the Gemini app and Search, but access rules differ by product; Google's performance claims are not independent TideReach tests.
NVIDIA announced on 12 November that Jensen Huang had resigned immediately as president, CEO and a board member. Colette Kress becomes interim CEO while retaining her CFO role, putting new leadership over the company's AI and accelerated-computing business.
Hugging Face published a guide on 12 November for adapting TRELLIS.2, an image-to-3D AI model, to footwear geometry. It trains low-rank adapters in the model's structure and shape stages while keeping the original weights and texture stage frozen.
Google has begun bringing the Gemini app to Apple CarPlay, extending its conversational AI interface to compatible car displays. Users can talk with Gemini Live hands-free and continue their chats on an iPhone afterward. The rollout begins globally, but that does not make every connected-app or personalization feature …