Anthropic has released a system card for 'Claude Mythos,' a new preview model. The document provides a rare look into the company's internal safety testing.
AIBW Security DeskAnthropic's new Project Glasswing initiative is tackling AI safety from the ground up by formally verifying critical software that powerful models depend on.
AIBW Security DeskA growing number of developers are voicing frustration on GitHub and Hacker News, claiming recent updates have made Anthropic's Claude 'unusable' for complex coding tasks.
Builder LabYour 3-minute AI news briefing for Tuesday, April 7, 2026. The top stories in artificial intelligence, delivered daily.
Daily DigestA new analysis reveals ChatGPT uses a sophisticated Cloudflare script that reads the app's internal React state to verify you're human before you can even type.
AIBW Security DeskOpenAI has unveiled its new Safety Fellowship, a pilot program designed to support independent research into AI safety and alignment and develop a talent pipeline.
AIBW Security DeskIn a landmark international collaboration, the US Cybersecurity and Infrastructure Security Agency (CISA) and the UK's National Cyber Security Centre (NCSC) have jointly released the 'Secure AI Framew
AIBW Security DeskA paper published by researchers at the Stanford AI Lab has detailed a novel jailbreak technique named 'Semantic Doppelgänger'. This method circumvents the safety filters of leading Large Language Mod
AIBW Security DeskEmerging AI unicorn Chroma Weaver AI announced a significant security incident affecting its flagship image generation model, Spectrum v4. Attackers exploited a previously unknown vulnerability in the
AIBW Security DeskThe European Parliament has officially passed the AI Safety and Accountability Act (ASAA), a comprehensive piece of legislation aimed at increasing the transparency and security of high-risk AI system
AIBW Security DeskA new paper published by researchers at the Stanford AI Lab details a powerful new jailbreak technique named 'Contextual Camouflage'. This method bypasses the safety filters of major large language mo
AIBW Security DeskNexusAI, a leading provider of enterprise AI solutions, has confirmed a significant data breach affecting its flagship 'Cognito-7' model infrastructure. The incident, detected on February 15, 2026, re
AIBW Security Desk