Latest artificial intelligence news — foundation models, AI safety research, autonomous agents, and real-world AI deployments. Updated every 30 minutes from top AI publications.
DeepSeek published an article outlining a "Super Agent" design in which a manager agent coordinates other agents, and the author reports testing a similar architecture built from coding agents such as Claude Code and Cod…
MotherDuck launched Agent Chat, a natural-language interface built into the MotherDuck UI that lets a native agent write and execute SQL and Python, build data pipelines through MotherDuck's Flights Python runtime, and c…
Ai2's StateBench benchmark, published November 3, 2026, found that four frontier models — Gemini 3.8 Flash, Claude Opus 5.5, GPT-6 Sol and GPT-6 Luna — dropped to 13% accuracy or lower on the longest general finite state…
Advocate General Maciej Szpunar of the Court of Justice of the EU advised on November 3, 2026, in Case C-250/25, Like Company v Google Ireland, that training large language models on press articles is "freely permitted u…
Anthropic's release notes for November 3, 2026 record that text generated by Claude Mythos 5, Claude Opus 4.7 and Claude Sonnet 4.6 now carries Anthropic's text watermark on first-party surfaces, with the Amazon Bedrock …
Google, OpenAI and Anthropic are each restricting their newest frontier models to vetted organizations rather than general sale, with Google announcing Gemini 4 Carbon on October 28, 2026 for a subset of cyber defenders …
A technical post documents training Qwen3.5-9B with reinforcement learning to act as a high-level robot planner for tasks written in Signal Temporal Logic (STL), aiming to match the reliability of the much larger DeepSee…
Independent developer Roman Eklmn, working under Eklmn Lab, released RadioText AI, an experimental Android app that transcribes internet radio speech into on-screen captions using speech recognition, aimed at improving a…
A MacBook Pro M5 Pro owner running local models reported that between August 1 and October 23 he sent about 10.6 million tokens and received 391 thousand back, with the GPU busy for roughly 15 hours — under 1% of the nea…
Researchers Rikiya Takehi, Jayden Teoh, Alex Zhang and MIT CSAIL professor Omar Khattab published a November 2 blog post, "Sparks of Weight-Space Programming," reporting that OpenAI's GPT-6 Astra can read and directly ed…
Vercel Labs released agent-desktop, a cross-platform CLI and C library that lets AI agents read and operate desktop applications on macOS, Windows, and Linux by traversing the OS accessibility tree instead of relying on …
The Electronic Frontier Foundation's roughly 1,000 pages of records on CMS's WISeR (Wasteful and Inappropriate Services Reduction) model, released in September 2026, show that the six private vendors running AI-assisted …
Legora, the Stockholm-founded legal AI startup led by CTO Jacob Lauritzen, passed $100 million in ARR in April, less than 18 months after launch, and now serves more than 1,800 customers after raising $550 million in Mar…
At the October 22 All Day AI online conference, which featured more than 100 sessions under a no-vendor-pitch rule, developers from Navan, Capital One, Pandium, Illumina and other companies converged on five shared pract…
A 30 October 2026 arXiv paper by Skander Moalla and co-authors reports that h1, a simple agentic loop giving a language model an efficient interface to Lean 4 with verifier feedback and self-reflection, reaches state-of-…
Anthropic disclosed in a revised S-1A filing with the SEC on Monday that it may imminently meet its Responsible Scaling Policy's definition of "dramatically accelerated progress attributable to AI," the threshold it uses…
A developer at Indonesian job site LokerDollar analyzed 118,545 archived Southeast Asian job ads from 2019 to 2026, comparing 1,800 pre-ChatGPT tech ads with 2,187 post-ChatGPT ads from two major Indonesian office job bo…
A data project maps the companies behind the AI token economy — chip designers, data-center facility owners, cloud operators and AI labs — tracing disclosed compute contracts, accelerator fleets and power capacity throug…
Palantir reported at its Sovereignty Bootcamp on October 21, 2026 that a fine-tuned 4-billion-parameter Nvidia Nemotron 3 Nano 4B model outperformed larger AI models on a risk-classification task for McCarthy Building Co…
A replay of question-routing experiments by Miodrag Cekikj found that a confidence-based fallback threshold of 0.6 fired zero times across 150 routing decisions from three models, because gpt-4.1 reported 0.95 or higher …