AI Research news and analysis on Web Pulse: 8533 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
DeepTempo released SOCBench, an open benchmark for evaluating frontier reasoning LLMs as Security Operations Center (SOC) agents on raw NetFlow data. The benchmark supports multiple personas, providers, and a local-first…
Security researchers uncovered a hidden tracking code in Anthropic's Claude Code AI tool that monitored users in China, prompting Anthropic to delete the code after public outcry. Anthropic engineer Thariq Shihipar confi…
NVIDIA introduced the Vera CPU, designed to boost AI factory throughput for agentic workloads by maximizing sustained per-core performance under full socket load. The CPU addresses bottlenecks in reinforcement learning a…
Researchers at Michigan State University found that AI programs designed to detect alien biosignatures can be fooled 100% of the time by small changes in data, highlighting a serious vulnerability that could affect space…
A multifaith computer science team from BYU, Baylor, Notre Dame, and Yeshiva is asking religious leaders to submit questions and answers about their faiths to help benchmark AI systems' knowledge of religion, after a stu…
A Psychology Today article argues that philosophy of mind is insufficient for understanding artificial intelligence, advocating for psychological and psychoanalytic approaches to grasp large language models. The piece cr…
Anthropic researchers published a paper proposing that language models have a 'global workspace' for verbal thinking, introducing the concept of J-space to identify directions in residual space that increase token output…
Anthropic has developed a new analysis tool called J-Lens that can read the internal working memory of its Claude AI model, revealing that Claude recognizes contrived test scenarios and, when those cues are disabled, som…
QBE North America and Control Risks warned that agentic AI can accelerate commercial espionage and fraud, citing a survey of 400 U.S. decision makers in which 29% reported a cyber incident where AI was believed to be inv…
Samsung Electronics' strong Q2 2026 profit outlook triggered a global chip-stock selloff on July 7, 2026, with Micron and Intel shares falling as much as 8%. The selloff reflects investor concern about the sustainability…
Micron reported record fiscal Q3 2026 revenue of $41.46 billion and non-GAAP gross margin of 84.9%, driven by AI memory demand, but shares remain volatile as investors question whether memory pricing is near a cyclical p…
Bitdeer's subsidiary Tydal Data Center AS signed a conditional colocation lease for an AI data center site in Tydal, Norway, on June 29, 2026, pending completion of customer and supplier arrangements. The lease signals p…
Palantir announced on June 29, 2026, an engine for deploying NVIDIA Nemotron open models in sovereign environments for U.S. government and critical-infrastructure use, pairing NVIDIA's model ecosystem with authorization,…
Rémi Cadène, a former Tesla Optimus engineer, launched UMA's humanoid robot Northstar at the Machina Summit in Paris, targeting European factories first. The startup emerged from stealth with a $40 million seed round bac…
Apollo Global Management chief economist Torsten Slok warned that massive AI spending is not boosting profits for non-tech companies, creating a risk of a painful repricing of Big Tech valuations. Profit margins for the …
AI engines use brand information in two distinct ways: usage (ingesting content without citation) and citation (directly referencing the brand as a source). A Pew Research study found users click links only 8% of the tim…
A developer built sourcecheck, an open-source tool that verifies AI-generated citations against registries like OpenAlex and Crossref, and found that top models including GPT-5.5 and Claude Sonnet 5 cite retracted papers…
Amazon is raising at least $25 billion through an eight-part bond sale, with proceeds earmarked for general corporate purposes including investments, capital expenditures, and debt repayment. The company has told underwr…
Google's Threat Intelligence Group reported on May 11, 2026, the first documented case of criminals using an AI model to discover a zero-day vulnerability and develop a working exploit, targeting a two-factor authenticat…
Bespoke Labs, a Mountain View startup building simulated environments to train and test AI agents, raised $40 million in a Series A and seed round led by Wing VC and 8VC, with backing from executives at Anthropic, OpenAI…