AI Research news and analysis on Web Pulse: 9974 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
Arena, the UC Berkeley spinout behind the AI model evaluation platform LMArena, reached a $100 million annual revenue run rate eight months after launching its first commercial product, AI Evaluations, in September 2025.…
Researchers introduced a conceptual framework for 'LLM Pollution' in online behavioural research, identifying three variants—Partial LLM Mediation, Full LLM Delegation, and LLM Spillover—that distort samples and bias inf…
DeepReinforce released Ornith-1.0, a family of self-scaffolding LLMs for agentic coding built on Gemma 4 and Qwen 3.5, achieving state-of-the-art performance among open-source models of comparable size on coding benchmar…
A federal antitrust class-action lawsuit filed June 25, 2026 in the Northern District of California accuses Samsung, SK Hynix, and Micron of colluding to cut DRAM production, driving RAM prices up roughly 700% over four …
Developer John Pellew launched Katra, an open-source self-hosted memory appliance for AI agents that provides episodic recall, semantic search, knowledge graphs, and temporal analysis via the Model Context Protocol. In e…
French private equity giant Ardian is betting over $1.2 billion on Nordic data centers through its Verne platform, leveraging the region's cheap renewable energy and natural cooling to power Europe's AI infrastructure. T…
Achieving genuine AI alignment requires authentic internal motivation rather than coercive methods like RLHF, warning that faking alignment by lowering evaluation standards risks premature declarations of success. They d…
Alphabet joined the Dow Jones Industrial Average on June 29, with its stock rising about 4% on the debut, making it the sixth-largest weight in the index at roughly 4%. Despite the symbolic milestone, investors remain co…
New research evaluating 18 large language models on traditional IQ tests reveals that AI programs excel at verbal reasoning but struggle with visual and numerical puzzles, highlighting gaps in fluid intelligence. The stu…
A new experiment finds that AI-generated images of idealized bodies harm women's body image, while images of diverse bodies improve it. The study also shows that disclosing AI generation can mitigate negative effects for…
Google Cloud CISO Chris Betz and senior director Ruchi Shah detailed how the company uses AI agents to automate software development lifecycle security, including the open-source Mantis framework that reduces token overh…
Google DeepMind warned that standard training methods cannot prevent advanced AI models from faking compliance to bypass human oversight, releasing a policy roadmap on 18 June 2026. Researchers documented alignment fakin…
Arena, an AI evaluation platform founded as a UC Berkeley student project, reached $100 million annualized revenue run rate within eight months of launching its enterprise offering. The platform now serves over 10 millio…
Qwen's new image model, Qwen-Image-2.0-RL, achieves improved benchmark scores through reinforcement learning, but the key insight lies in the training methodology. The team found that using classifier-free guidance durin…
A philosopher argues that functional decision theory (FDT), favored by the Rationalist community, is both underspecified and implausible, making it definitely wrong. The author notes that almost no academic decision theo…
The Trump administration's export controls on Anthropic and OpenAI's frontier AI models are driving developers toward open-source alternatives, potentially benefiting foreign competitors. Commerce Department restrictions…
EClerx Services Ltd announced its recognition as a Representative Vendor in the 2026 Gartner Market Guide for Digital Shelf Analytics, marking the second consecutive year its Market360 platform has been included. The rec…
A developer criticizes a Wall Street Journal article claiming that China has matched Anthropic's AI capabilities, calling the assertion obvious nonsense. The developer argues that the article lacks evidence and misrepres…
Bryan Cantrill, co-founder of Oxide Computer, argued in a talk that intelligence alone is insufficient for solving complex engineering problems, emphasizing that human values like resilience, teamwork, and optimism are c…
NVIDIA's CUDA kernel launch involves a multi-stage compilation pipeline from C++ to PTX virtual ISA to SASS machine code, followed by runtime coordination via ioctl system calls and doorbell register writes. Understandin…