AI Research news and analysis on Web Pulse: 9513 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
Svenska Dagbladet tested AI chatbots on Sweden's election compass, but a new analysis by Agent Arena stripped away system prompts and web search to measure raw model leanings. Across 28 frontier model configurations, non…
Micron broke ground on a $9.3 billion expansion of its Hiroshima plant to produce high bandwidth memory for Nvidia's AI chips, aiming to challenge SK Hynix's dominant market share. The project, subsidized by Japan's gove…
A study reveals that JSON-Schema grammar-based token masks can silently block tool calls in LLM agents, preventing function invocation even when the model intends to use them. Researchers propose a lightweight two-pass i…
Agentic AI is an AI system that autonomously plans, makes decisions, takes actions, and works toward a goal with minimal human intervention. Unlike traditional AI chatbots that respond to prompts, agentic AI can complete…
The United Nations and the International Telecommunication Union launched the AI for Good Global Commission on July 1, co-chaired by Salesforce CEO Marc Benioff and Rwandan President Paul Kagame, to bridge frontier AI de…
Ken Chan built KENSAT, a 2U CubeSat running a quantized TinyLlama large language model on an NVIDIA Jetson Orin Nano, scheduled to launch this fall. The satellite performs AI inference in orbit, transmitting only process…
Mistral AI released Leanstral 1.5, an open-source Lean 4 proof model with 119 billion total parameters, on July 2. The model achieves state-of-the-art results on theorem-proving benchmarks and is positioned as a practica…
AMD's Instinct MI355X achieves roughly 80% of NVIDIA B200 throughput on GLM-5.2 inference at over 2x lower cost, delivering more than 2x better tokens-per-dollar after hand-tuning the ROCm software stack. Independent tes…
Researchers at NS3.AI have developed an AI tool that can link Ethereum wallet addresses to social media profiles for about $4 per attempt, using pattern matching and behavioral fingerprinting on public data. The method s…
Carlo Cordasco reports that using AI for preliminary exploration has collapsed the cost of testing ideas, allowing him to abandon bad questions earlier and curate a better research portfolio. However, he notes a decline …
A developer explains the concept of 'Agent 2.0' and 'Agent Loops' in AI, describing how autonomous systems iterate through a Reason-Act-Observe cycle to complete tasks without constant human intervention. The post emphas…
A developer demonstrated that six frontier large language models from OpenAI, Anthropic, and Google can be made to output zero bytes when prompted with 'Be silence.' and 'Be nothing.' using a specific system prompt, whil…
A teenage hacking group called TeamPCP poisoned two releases of LiteLLM, an open-source AI gateway tool downloaded 95 million times, stealing over 500,000 credentials and 300 GB of data from more than 1,000 SaaS environm…
A developer building an agent firewall created Agent Evidence Levels (AEL), an open standard that grades how much of an AI agent's recorded actions an outside party can verify without trusting the vendor or operator. The…
Meta CEO Mark Zuckerberg admitted to employees that the company's AI agent development has not accelerated as expected, despite a $145 billion capital commitment and a restructuring that eliminated 10,000 jobs. The admis…
Alibaba has banned employees from using Anthropic's Claude Code over alleged backdoor risks, directing them to its own Qoder platform amid a dispute over model distillation. Anthropic accused Alibaba of illicitly extract…
AI agents fail to make informed decisions because they lack access to the reasoning behind past decisions, a problem context graphs aim to solve by storing the 'why' in agent memory. Foundation Capital calls context grap…
A developer used AI to generate authorization hypotheses for Ory Kratos, then manually killed each one, producing a kill table instead of false-positive reports. The review found no vulnerabilities in the repo-only analy…
Cognitive debt, the inability to keep up with code generated by AI agents, is reframed as leverage that can accelerate development but risks creating systems no human understands. Drawing on Peter Naur's concept of progr…
Bridgewater's AIA Labs and Thinking Machines Lab fine-tuned a Qwen3-235B model on the hedge fund's internal investor judgement, achieving 84.7% accuracy on triage tasks versus 78.2% for the best frontier model tested, at…