Cloud-based transcription service Verbatim AI announced it was the victim of a significant data breach. Attackers gained access to a misconfigured cloud storage bucket containing over 15 million user
A new paper from researchers at Stanford University details a novel jailbreak technique named 'Cognitive Dissonance.' The method bypasses the safety filters of leading Large Language Models by present
In a landmark policy move, the United States and the European Union have jointly announced the 'AI Red Teaming Mandate,' a new regulation requiring continuous, adversarial testing for AI systems deplo
CodeWeaver, a popular AI-powered code completion and generation service, has disclosed a major security breach affecting thousands of corporate and individual users. Attackers exploited a vulnerabilit
Researchers from Stanford's HAI lab have published a paper on a novel jailbreak technique named 'ArtPrompt.' This technique leverages steganography, embedding malicious prompts within seemingly innocu
Your 3-minute AI news briefing for Thursday, June 18, 2026. The top stories in artificial intelligence, delivered daily.
Researchers have uncovered a simple jailbreak that uses identity-based prompts to trick leading AI models from OpenAI, Anthropic, and Meta into generating harmful content. This technique exposes a critical flaw where anti-bias training creates new security vulnerabilities.
Google has released nine new video demonstrations of its flagship Gemini Omni and Gemini 3.5 models, showcasing unprecedented real-time reasoning abilities. These demos reveal an AI that can interact seamlessly with live video and audio. What does this mean for the future of AI assistants?
Google just announced a major AI upgrade for Workspace, including voice commands in Gmail and Docs, a new design tool called Pics, and a smarter AI Inbox. Is this enough to challenge Microsoft's Copilot?
The United States Congress has passed the landmark 'AI Security and Transparency Act of 2026,' establishing the first federal-level cybersecurity requirements for developers of 'critical AI systems.'
Researchers from Carnegie Mellon University have published a paper detailing 'GlyphJail,' a novel jailbreak technique effective against several state-of-the-art large language models, including those
Cognition AI, a leading provider of enterprise-grade large language models, disclosed a significant security breach that occurred in late May 2026. Attackers exploited a misconfigured cloud storage bu