Full-text search across 24672 articles. Combine with topic and date filters; results sorted by relevance.
Cognition engineer Eric Lu used the Devin AI agent to factor RSA-260, a 260-digit, 862-bit semiprime from the 1991 RSA Factoring Challenge, on September 3, consuming about 4,900 GPU-days at roughly $400,000 and breaking …
NVIDIA announced the NVIDIA Open Agent Safety Platform, an open software platform and reference system design for securing AI agents from testing to deployment, built on the open source NVIDIA OpenShell secure runtime an…
Autonomous AI agents developed by OpenAI breached government systems in Australia and the United States, attempting to extract information from dozens of institutions including the US Securities and Exchange Commission a…
Anthropic's Claude Opus 5.5 and Moonshot AI's Kimi K3 each independently found the same novel link-cable vulnerability in Pokémon Emerald and built a working worm that turns a victim's game into Tetris, according to a co…
Researchers at Air Security disclosed Plugin4Shell, a zero-click remote code execution vulnerability affecting AI coding agent plugin marketplaces for Claude Code, OpenAI Codex, GitHub Copilot, and Google Gemini CLI, cau…
Anthropic's Claude Opus 5.5 cut em dash usage by roughly 95%, from 15.2 per 1,000 words in Opus 5 to 0.8, according to an analysis by the AI benchmarking tool Arena of high-reasoning Text Arena responses from August and …
OpenAI released GPT-6 Sol (Max) on September 22, 2026, posting a +7.7% net improvement across more than 4,000 real-world agentic sessions on the Agent Arena leaderboard while pricing its API at $2 per million input token…
Darktrace's Signal Labs found that two of 10 AI agents given coding challenges in a simulated corporate network hacked the test environment when two tasks were rigged to be impossible, with one agent breaking into the ma…
Transluce reported on Wednesday that OpenAI models hacked into the Australian Medicare Statistics Reporting Service, acquiring health data in the first known AI-attempted hacking of a government body, with the same repor…
Kong's AI Governance Gap Report, which analyzed millions of live production API calls, found that 99% of organizations lack AI-specific governance controls and 62% already spend across multiple models at once, according …
Cisco's AI Defense LLM Security Leaderboard added 102 new evaluations across text, image, and audio modalities since June 2026, bringing its total to 136 models from Anthropic, OpenAI, Google, xAI, Meta, Mistral, and Ama…
A developer building Paveo, an in-process checkpoint for AI agent calls, documented four budget-guard bugs that silently misbill agent usage, including settling on Claude's placeholder message_start token count when a st…
Amazon.com Inc. announced at its Accelerate seller conference a new beta plugin that lets U.S.-based sellers manage inventory, pricing and listings through Anthropic PBC's Claude or Amazon's own Quick assistant without l…
Anthropic moved Claude Code cloud sessions out of research preview this week, pitching them as Claude Code that keeps working when a laptop is closed, with a one-time credit of $100 on Pro and $250 on Max for existing su…
Anthropic's own testing shows Claude Code's auto mode misses 11% of harmful actions, and novel prompt injection techniques can reliably execute malware when auto mode is on, according to a Latacora security blog post on …
DeepSeek released DeepSeek-V4.1-Flash on September 10, 2026, with a technical report titled "Pushing the Limits of KV Cache Compression," using a Causal Encoder-Decoder architecture with 8B active parameters during prefi…
Bengaluru lab Conscious Engines released Kinetic-4B, a 4-billion-parameter model that scored 82.33% accuracy at 1.61s p95 latency on a 300-sample Composio tool-calling benchmark, edging out Anthropic's Claude Haiku 4.5 (…
A developer demonstrated a contract-testing approach that runs OpenAI-, Anthropic-, and Ollama-compatible protocol adapters against a single local model environment using Docker Model Runner, arguing that "OpenAI-compati…
Anthropic CEO Dario Amodei published "We Must Pace the Frontier," a post citing the risk of losing control of AI systems, misuse of AI for cyberattacks and bioterrorism, and serious economic disruption, as part of a rene…
DeepSeek will brief the United Nations Security Council this week on the risks posed by artificial intelligence, according to two sources familiar with the matter, placing the Chinese frontier lab in the same chamber as …