AI News, 14 November 2026

Saturday 14 November

The Pentagon is testing war robots that can hunt in packs without GPS and find battlefield targets on their own. Humans still call the shots.

AI-generated summary The Defense Advanced Research Projects Agency tested autonomous uncrewed ground vehicles that conducted surveillance and searched for targets without GPS at Fort Knox, Kentucky. Jason Gregory says human operators maintained control and situational awareness throughout the testing of artificial intelligence algorithms and ground mobility under the Robotic Autonomy in Complex Environments with Resiliency program.

Google Cloud SC26 Agentic HPC and AI Clusters

Google Cloud SC26 Agentic HPC and AI Clusters

AI-generated summary Google Cloud announced updates to its high-performance computing and artificial intelligence infrastructure stack ahead of the Supercomputing 2026 conference. The release introduces composable agentic high-performance computing workflows and DeepMind Science Skills designed to automate infrastructure orchestration and validation.

Earlier:

Claude Code is quietly becoming a hardware hacker's best helper, and Linux support is next

AI-generated summary XDA reports that users are applying Claude Code to build custom smart home sensors and program microcontrollers. Christian Schaller used Claude Code to help develop a Linux driver for a Plustek OpticFilm 8200i scanner and fix a webcam issue on a ThinkPad X1 Carbon Gen 14.

Earlier:

Avanade Agrees to Buy Robiquity for UK AI Automation

Avanade Agrees to Buy Robiquity for UK AI Automation

AI-generated summary Avanade agreed on November 10, 2026, to acquire Robiquity to add UK delivery capability in artificial intelligence, intelligent automation, and low-code development. Growth Capital Partners has signed an agreement to realize its investment in Robiquity, with financial terms of the transaction remaining undisclosed.

Earlier:

MIT FLEX Framework: Robot Nonliteral Commands

MIT FLEX Framework: Robot Nonliteral Commands

AI-generated summary On November 9, 2026, researchers from MIT and New York University introduced FLEX, an AI-assisted robot learning pipeline designed to interpret nonliteral human instructions during physical motion planning. The framework queries OpenAI’s GPT-4o to analyze registered features alongside prior constraints, allowing the language model to infer latent abstractions that human operators did not explicitly verbalize.

Earlier:

I used Claude Code to build the browser feature I've been wanting for years

AI-generated summary The author used Claude Code to build a Chrome extension that replicates the Glance feature from the Zen browser. Using Opus 5.5 in web mode, the tool generated a working extension with customizable shortcuts and floating preview windows in about twenty minutes.

Also covered byAgentScout
Earlier:

Microsoft Winds Down Oracle AI Cloud Deal As Older NVIDIA GPUs No Longer Meet Its Needs

AI-generated summary Microsoft has reportedly decided to wind down its artificial intelligence computing agreement with Oracle because older NVIDIA graphical processing units no longer fit its evolving technical requirements. Oracle maintains that demand for its previous-generation accelerators remains robust, reporting high utilization and pointing to recent contract renewals at a premium.

Earlier:

Claude Fable 5.2 vs. GPT-6 Astra: A lower list price didn’t mean a lower bill

AI-generated summary Anthropic launched Claude Fable 5.2 on Nov. 9 with standard API rates of $8 per million uncached input tokens and $40 per million output tokens. OpenAI launched GPT-6 Astra in September with standard API rates of $10 per million uncached input tokens and $50 per million output tokens.

Also covered byCloudNinjasBytechap
Earlier:

LangChain’s $3 Billion Valuation Puts AI Agent Control at the Center

AI-generated summary LangChain has raised $300 million at a $3 billion valuation in a funding round led by CapitalG with participation from GIC, DFJ Growth, IVP, and Sapphire Ventures. LangChain reports having 7,000 LangSmith customers, including Harvey, Rippling, Clay, Nvidia, Toyota, and Salesforce, as it focuses on the creation, deployment, and management of AI agents.

Also covered byTopFiveTwarxFathom News
Earlier:

Meet the seven co-founders of Anthropic

AI-generated summary Four researchers, a former journalist, a public policymaker, and an engineer founded the artificial intelligence company Anthropic in 2021. Trading for the public offering is expected to begin on Nov. 19, and the seven co-founders collectively own about 10% of the company.

Earlier:

Lawyer upgrades eight-V100 private AI server to run 80% of DeepSeek V4.1 Flash at 4-bit NVFP4 precision — adds 1TB of RAM, an Nvidia RTX 3090, and fixes crashes caused by a bad PSU

AI-generated summary A lawyer upgraded a private AI server built on eight used Nvidia V100 GPUs to run about 80% of DeepSeek V4.1 Flash at 4-bit NVFP4 precision. The system also received 1TB of RAM and an Nvidia RTX 3090, while a bad power supply was identified as the culprit behind previous crashes.

Also covered bymlllm.io
Earlier:

What Is the NIST AI Risk Management Framework? A Guide for AI Teams

AI-generated summary The NIST AI Risk Management Framework is a voluntary, use-case-agnostic framework designed for incorporating trustworthiness and risk management across the artificial intelligence lifecycle. Unite.AI explains that the concept includes practical commitments such as an identifiable input, a characteristic transformation or decision, and an outcome that can be evaluated against a stated objective.

AWS Announces New AI Tools & Models for Builders

AI in Practice

AI-generated summary AWS announced updates to Amazon Bedrock and AgentCore to provide AI builders with greater flexibility, lower costs, and improved management tools. The updates are designed to empower AI developers, machine learning engineers, and data engineers to choose optimal models and integrate AI into core workflows.

Earlier: