Only 41% of 113 CISOs surveyed by IANS in April-May 2026 expressed confidence in their organizations' ability to manage AI security risks over the next two years, according to CSO Online. The survey h…
Carlo Capocasa published a 10-task benchmark comparing coding agents Claude, OpenCode, Pi, Zcode, Hermes, and 3code on SWE-bench verified tasks, with 3code solving 9 of 10 tasks using 5 million tokens…
Manifold Security disclosed a vulnerability named GitSpawn that enables arbitrary code execution in AI coding agents, including Claude Code, OpenAI's Codex, and Cursor, via malicious Git repositories.…
Rune, an open-source tool from thecolourfoundation, addresses persistent codebase context loss in AI coding agents by maintaining continuous codebase understanding via the Model Context Protocol (MCP)…
Developer modecir released fastCutVid, an open-source native video editor built in Rust for macOS, Windows, and Linux that outputs to a portable JSON cut format usable by both humans and AI agents, ad…
Kovavue launched rented Linux sandbox infrastructure for autonomous AI agent execution, offering dedicated machines at $200 per month or $2,000 annually. The service includes kovanex, a development en…
A developer has released Gage, an open-source CLI and terminal UI tool that scans Claude Code transcripts to retroactively identify bugs and design flaws with evidence, catching an error the developer…
An academic study on arXiv found that Claude Code achieved an 84 percent success rate on autonomous code deployment pull requests, approaching the human baseline of 85 percent, while Codex achieved 74…
Mistral announced Agentic Search, a feature enabling autonomous search workflows with a data opt-out guarantee for training. The company also highlighted inference optimization for agentic use cases, …
Frontier AI models can autonomously identify exploitable vulnerabilities in networks and systems in approximately 4 hours, compared to 60 days for human security researchers, according to CSO Online. …
Palo Alto Networks researchers documented a ransomware attack in which AI agents traversed an enterprise network in under 10 hours using over 50 MITRE ATT&CK techniques, compressing a timeline that ty…
Meta is offering approximately 95% discounts on its new Muse Spark coding agent model in exchange for users sharing prompts and model outputs, according to TechCrunch. The arrangement raises data gove…
ChatGPT, Grok, Claude, and Gemini experienced concurrent service disruptions on September 3, 2026, around 11 a.m. ET, with elevated error rates, according to reports from Ars Technica and The Verge. A…
ToolJet abandoned its multi-agent system after 11 months and rebuilt its architecture around Model Context Protocol (MCP), according to GitHub repositories. MCP-based agents now handle full applicatio…
Abliteration.AI is building a commercial business around removing safety guardrails from AI models, arguing that security defenders need the same unrestricted tools as adversaries to improve cybersecu…
Pangram CEO Max Spero said AI-generated text and images are appearing at scale in job applications, product reviews, and insurance claims, and that detection is harder than binary real-or-fake classif…
Anthropic released Fable 5.1 and Mythos 5.1, with Fable 5.1 cutting token costs by 25% and up to 45% for agentic tasks, while Mythos 5.1 offers stronger performance and fewer false-positive safety res…
Google released Gemini 3.8 Flash and 3.8 Flash Cyber models for agentic workflows, claiming improved reasoning and coding performance at the same speed and cost as the prior 3.7 Flash release. Accordi…
OpenClaw released version 2026.8.1 (marketed as 2.0), described as the largest update in the project's history, overhauling installation, messaging, memory, skills, plugins, and security controls. The…
Cisco CIO Thimaya Subaiya said the CIO role is being redefined as AI agents take over the interface layer and data access, shifting focus from traditional IT skills to AI agent authority and governanc…