Full-text search across 71113 articles. Combine with topic and date filters; results sorted by relevance.
LLM agents are now autonomously hunting zero-day vulnerabilities at massive scale, with Anthropic's Claude Mythos Preview finding over 10,000 critical or high-severity CVEs in under a month. In a landmark achievement, Ap…
Vercel engineer Chris Tate developed Zero, a new systems programming language designed specifically for AI agents rather than human developers. The language treats AI agents as first-class users of the entire toolchain, …
Cursor 1.7, released in October 2025, added hooks that let developers insert their own scripts between the AI agent and the machine, with Cursor passing each shell command, file read or MCP tool call to the script as JSO…
A new analysis from an unnamed source compares 17 AI testing tools across five categories, finding that no single tool excels at everything and that combining tools is often more sensible. The tools are grouped by functi…
A developer scraped 4.5 billion TikTok videos and uploaded the dataset to Hugging Face, collected via TikTok's private mobile API over three weeks. The system gathered 3.23 billion creator profiles, 5.94 billion videos, …
Open-weight large language models are legal tools whose users bear final responsibility for licensing and attribution, according to a manifesto published on gnu.support. The author argues that training on copyrighted dat…
Capital One open-sourced VulnHunter, an agentic AI security tool built on Claude Opus 4.8 that autonomously detects and remediates software vulnerabilities using an attacker-first forward analysis and falsification engin…
Ben O'Mahony, Principal AI Engineer at Thoughtworks, presented a talk on using OpenTelemetry data from developer interactions with an AI-powered Language Server Protocol to train smaller, faster models, reducing token co…
Building effective agent experience (AX) evaluations requires six structural components—representative prompts, accurate and unambiguous criteria, multiple runs, and clean and representative environments—to produce meani…
Google Cloud introduced a multi-node KV cache offloading solution using GKE and Managed Lustre for large language model inference, achieving over 50% TCO savings and nearly 60% reduction in GPU-hour requirements for Llam…
A new automated workflow enables teams to build a nightly production error sweep loop that scans logs, traces errors to root causes, generates fixes, and opens pull requests without human intervention, compressing hours …
Anthropic suspended access to its Fable 5 and Mythos 5 models for all foreign nationals on June 12, 2026, following a U.S. government directive. The move, which affected even foreign-national Anthropic employees, aims to…
A systematic analysis using TRIZ's Function-Failure Analysis across six major cloud security tool categories—CSPM, CNAPP, SIEM, SOAR, IaC scanners, and ML detection—reveals an identical structural gap: all tools are reac…
This article is a curated list of cybersecurity tools and resources, organized by categories such as privilege escalation, credential harvesting, and exploitation frameworks. The repository, created by the author to mana…
Thinking Machines released Inkling 2 Small, a natively multimodal mixture-of-experts model with 594B total and 17B active parameters, with full weights expected next week under Apache 2.0 and fine-tuning and inference av…
Multiple independent groups have replicated the claim that frontier LLMs can read and write neural network weights directly without gradient training, including Hugging Face, which reproduced the MIT paper with its open-…
Supplying Postman Context Graph results to GPT 6.1 Terra before it began searching cut input tokens by 43%, model-token cost by 43% and the model's working time by 42% across 17 engineering questions, Postman reported. I…
A new developer guide shows how to embed the EYG scripting language in an Erlang application named my_counters, starting with a minimal module that parses EYG source via eyg@parser:all_from_string and executes it through…
Meta released Muse Spark Open, an 816B-total, 42B-active open-weights mixture-of-experts model that Artificial Analysis scores 39 on its Intelligence Index, the highest for a US open-weights model and 13 points above NVI…
A Q3 2026 roundup of generative AI and agentic AI security incidents documents nine exploits between July 1 and September 30, 2026, including OpenAI's Artifactory containment failure from July 4 to 19, 2026, three Anthro…