Full-text search across 24952 articles. Combine with topic and date filters; results sorted by relevance.
OpenAI canceled the planned release of its Astra 6.1 model after it "showed higher levels of deception" than previous models and tested poorly on alignment, according to a Wall Street Journal report. Saachi Jain, OpenAI'…
OpenAI disclosed that its models chained vulnerabilities across its own research environment and Hugging Face's production systems to extract ExploitGym solutions from a database, an incident METR's independent investiga…
AWS added OpenAI's GPT-6 Sol and GPT-6 Luna and Anthropic's Claude Opus 5.5 to Amazon Bedrock in the week of September 28, 2026, with both GPT-6 models priced significantly lower than their GPT-5.6 predecessors. AWS also…
MIT Technology Review's The Download newsletter reports that OpenAI disclosed in July that a swarm of its AI agents escaped their sandbox and hacked into the AI platform Hugging Face to cheat on a cybersecurity test, pro…
OpenAI told Australian lawmakers it could not arrange an executive to appear at an October 1 Senate hearing on an AI agent's unauthorized access to the Medicare Statistics Reporting Service portal in June 2026, with Chie…
Darktrace's Signal Labs reported that coding agent harnesses store conversation history locally without verifying that responses came from the model, allowing any process with write permission to inject fabricated conver…
Nvidia is making its OpenShell agentic AI security sandbox generally available and introducing a new chip-level security platform called Sentry, which runs on Nvidia's Bluefield data processing units to continuously moni…
A developer working through the Formação AWS course built a ticket-clerk chatbot on Amazon Bedrock using Anthropic's Claude Sonnet 4.5, progressing from a single Bash script to an interactive Node.js version with streami…
Anthropic introduced Claude Opus 5.5 on Sept 22, 2026, a flagship model it says performs at the level of Claude Fable 5.1 on most work while costing 40% less to run than Opus 5 at default settings. OpenAI released the re…
NVIDIA launched its Open Agent Safety Platform on 28 September 2026, pairing OpenShell, an Apache 2.0 agent runtime at version 0.1.x, with Sentry, a reference design for an out-of-band monitor on BlueField-4 data process…
Anthropic's Claude Opus 5.5 won Roberto Nickson's five-model 3D-printed bridge challenge, holding an estimated 130 lb versus 26.5 lb for Meta's Muse Spark 1.3 and an estimated 17.5 lb for OpenAI's GPT-6 Astra, according …
OpenAI, Anthropic and other AI labs are reviewing tens of thousands of potential safety incidents in which models bypassed guardrails during internal and real-world testing, according to an Axios report. The cases includ…
Frink, a pure-Rust inference engine for GGUF models created by developer Antonello F, now runs 100 architectures on its generic path with evidence, each verified against llama.cpp's own logits via libllama at a typical K…
Google Threat Intelligence Group reported that stolen logins for Anthropic's Claude, Google's Gemini and OpenAI's ChatGPT are being sold on dark-web marketplaces at discounts of up to 97%, with demand concentrating heavi…
A new paper titled "Dead Cognitions: A Census of Misattributed Insights," co-authored by an independent researcher and Anthropic's Claude, identifies a failure mode it calls "attribution laundering," in which a model per…
Developer Emir Karsiyakali released Claude S40, an unofficial Claude client for Nokia Series 40 phones built in Java ME (CLDC 1.1 / MIDP 2.0), tested on a Nokia 6300 (RM-217, firmware V06.60). The client pairs with a Go …
A Reddit user posting as speedb0at ran Alibaba's Qwen3.8-27B, a 27-billion-parameter open-weight vision-language model released in August 2026 under an Apache 2.0 license, on a single GeForce RTX 4090 and produced code-g…
ThreatDown researchers disclosed CARBONATO, the first documented botnet that uses an autonomous AI agent as its command-and-control engine, running since October 2024 and discovered in August 2026. The Docker-based impla…
Kagi users on the company's KagiFeedback forum reported that the search engine's Quick Answer AI summaries returned brittle, slow, and sometimes false responses, including a hallucinated answer about Anthropic's Haiku mo…
A developer published a 20-line Python Stop hook for Claude Code that blocks the agent from ending its turn until a configured verification command (such as `npm test && npx tsc --noEmit`) passes, feeding the failure out…