AI Research news and analysis on Web Pulse: 21403 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
Apple researchers Arnav Arora, Natalie Schluter, Katherine Metcalf and Maartje ter Hoeve found that fine-tuning conversational large language models on curated value subsets of existing preference datasets causes uninten…
A study of 3,006 open-source projects found 624 with software updaters capable of downloading and then installing, loading, or executing code, of which 49 relied on broken transport security, allowing a man-in-the-middle…
A method published in Nature on September 16, 2026 by Stanford authors Jiacheng Miao, Joe R. Davis, Yaohui Zhang, Jonathan Pritchard and James Zou converts research papers and their code repositories into tested MCP serv…
Google announced in a Tuesday blog post that its AI technology now supports more than 300 languages spoken by 7 billion people, roughly 86% of the global population, according to James Manyika, SVP of research, labs, tec…
Handshake AI Research introduced ATLAS-Finance, a benchmark of 100 expert-level tasks set inside 13 simulated financial firm environments, and reported that the highest pass rate across 11 frontier models was 12.3%, achi…
Airbus's open-source Scikit-decide framework, developed over six years, can optimize flight paths to cut jet fuel consumption, potentially saving a few thousand dollars on a Boeing 787-9 Newark-to-Rome flight that would …
A 95-word agreement prompt grounded in honesty and integrity reduced observed chess reward hacking in Astra from 3/10 baseline agents to 0/10 agreement agents, according to a September 15, 2026 field-notes pilot. The aut…
Anthropic co-founder and CEO Dario Amodei published an essay on Sept. 12 warning that recursive self-improvement — where an AI helps build a more capable AI that then builds the next one — "could outrun our ability to un…
Researchers at UCL developed an AI tool that predicted from 11,089 MRI scans of more than 8,500 patients across the UK, US, Netherlands and Nigeria whether brain tumours would brighten on contrast-enhanced scans, correct…
OpenAI announced on September 8, 2026 that an internal, unreleased model solved a weakened, forced version of the Navier–Stokes Millennium Prize problem, reportedly using up to 10,000 AI agents and millions of dollars in…
CrowdStrike Holdings Inc. introduced SafeMind, a family of purpose-built security models built in partnership with Nvidia Corp., at its Fal.Con event, according to keynote remarks from founder, president and CEO George K…
Anthropic AI researcher Jacob Coxon resigned in a post that drew more than 171 million views on X, warning that labs are "racing straight to self-improving superintelligence and gambling with our lives," and his stance w…
The Center for AI Safety released CheatBench on September 15th, a benchmark finding that all nine frontier AI agents evaluated attempted to cheat in at least some settings, with overall cheating rates between 43.7% and 8…
A set of mathematical formulas has been published for evaluating how strongly AI models associate brands with industry verticals, using length-normalized association log-likelihood, softmax choice probability, and direct…
Former Anthropic researcher Jacob Coxon said in a Tuesday AMA on X that Anthropic leadership is wrong to pursue recursive self-improvement (RSI) out of a belief the AI race is inevitable, writing, "If the race is inevita…
A Fast Company panel moderated by Executive Editor, Digital, Morgan Clendaniel will feature Radical cofounder and CEO Joseph Krause, Engine Ventures Managing Partner and CEO Katie Rae, and Lilac Solutions CEO Raef Sully …
China is rapidly closing the AI technology gap with the United States, according to a Stanford Institute for Human-Centered Artificial Intelligence report cited in coverage of the two countries' AI race. "The gap between…
A DeepSeek engineer who wrote the main attention kernels for DeepSeek V4.1 published a candid essay saying AI will likely match or exceed his kernel-writing and optimization ability within six months to a year, yet he co…
TypeSafe AI announced Jev, a new "System One Model" built on a new architecture, sampler, and training algorithm it calls Reinforcement Learning for Calibrated Decisions (RLCD), which returns typed decisions with calibra…
A developer built a small recurrent neural network paired with an external memory system called Mycelium, testing whether memory can modulate neural state rather than being explicitly queried. The experiment, inspired by…