Full-text search across 17567 articles. Combine with topic and date filters; results sorted by relevance.
Monthly CVE disclosures doubled in 2026 to 10,740 by August, with high-risk vulnerabilities up 241%, according to Google Threat Intelligence Group data, and agent orchestration middleware now accounts for 50% of all AI-r…
A fork of Mert Cobanov's open-source PS5LM project, called PS5LM Plus, now runs as a native application on jailbroken PS5 consoles, freeing 11.4 gigabytes of memory versus roughly 6 gigabytes for the earlier payload vers…
Meta will roll out an updated commercial terms of service on Oct. 30 that authorizes the company "or any entity acting on our behalf" to use web scrapers, bots, crawlers and user agents to access, index, cache and crawl …
Microsoft's Azure AI Foundry launched a dedicated AI Model Marketplace that lets developers discover and integrate third-party AI models, alongside a new AI Governance Dashboard for monitoring model compliance and securi…
Microsoft Corp.'s AI research group detailed MAI-Base-2, a 2-trillion-parameter base model with a mixture-of-experts architecture that activates 80 billion parameters per prompt and, Microsoft says, outperforms DeepSeek-…
Agent Cyber Sentinel launched as a zero-dependency runtime security shield for autonomous AI agents and MCP servers, claiming under 1ms overhead and enforcement of the POD-9 (Proof-of-Defense in Nine Layers) specificatio…
Hands-on benchmarks of the MSI Prestige N16 Flip AI Plus, a 16-inch Windows-on-Arm laptop built on Nvidia's RTX Spark platform with a 20-core N1X chip and 128GB of LPDDR5 memory, found it outperforming Nvidia's DGX Spark…
Meta's new large language model Muse Wonder scored 79% on the Deep Suite coding benchmark, reportedly the top mark among frontier models, and ranked above Gemini 4 Argon, GPT 6 Astra and Claude Fable 5.1 on the Artificia…
Meta released the weights for its Muse Spark Open model today, with Mark Zuckerberg announcing on X that it runs efficiently on new Mac Studios through work with Apple's John Ternus. The M5 Ultra Mac Studio with 512 GB R…
Reflection AI Inc. released the weights of Beam, its first large language model, on Hugging Face, along with post-training recipes, a drafter model for speculative decoding, and ready-to-use configurations for several se…
A developer built an ambient expense-approval agent on Google's Gemini Enterprise Agent Platform using ADK 2.0 graph workflows, combining deterministic Python code nodes for parsing, thresholds, injection checks and reda…
Veeam Software Group GmbH announced Veeam Intelligent ResOps (iResOps) at its VeeamON Global Launch virtual event, with Microsoft 365 as the first generally available workload, arguing that security and data resilience m…
Elorian released Glimpse on October 16, its first public model, which the company says beats OpenAI's GPT-6 Astra on scientific chart reasoning and short-form video captioning while running at $0.60 per million input tok…
The .txt team released Outlines, an open-source library that guarantees structured LLM outputs during generation rather than fixing them afterward with parsing or regex, and it is trusted by NVIDIA, Cohere, HuggingFace a…
A developer built DukaanVoice AI, a local-first shop management and point-of-sale application for Indian kirana stores and small retailers, released as v1.0.1 on GitHub. The app, built with React, Vite, Python, FastAPI, …
Google's Gemini 4 Argon reached general availability on October 13, with introductory API pricing of $2 per million input tokens and $10 per million output tokens that doubles to $4/$20 on December 1, the same as Claude …
An Anthropic research model identified the origins of the NIST P-192, P-224, P-256, P-384 and P-521 elliptic curve seeds in under half an hour in October 2026, finding they were outputs of the C library's lrand48 random-…
Nvidia's GEAR lab released version 1.0 of ASPIRE on GitHub this month under an Apache 2.0 license, reporting 99% success across 80 tasks in the LIBERO-Pro benchmark and a robot arm solving a Rubik's Cube without task-spe…
SAIR soft-launched the SAIR Open Math OS, a web interface and service hosting existing open math models DeepSeek V4.1 Flash, Kimi K3, and GLM 5.3 Flash, which SAIR says are roughly a year behind current frontier models. …
A Benchmark Map analysis of 29 AI benchmarks against the US Department of Labor's O*NET taxonomy found that 90% of American working hours are spent on tasks with little or no coverage from existing AI benchmarks, with on…