A new project from a Hugging Face hackathon successfully runs a complex multi-agent economy on a tiny 3B parameter model. This breakthrough challenges the notion that only massive AI models can generate sophisticated, emergent behaviors.
Google is launching new AI-powered tools in Search and Shopping to revolutionize the second-hand market. These features allow users to find specific vintage and thrifted items using visual cues and descriptive language, tapping into a rapidly growing e-commerce sector.
The Open AI Security Foundation (OASF) has announced the first public release of 'Aegis,' an open-source framework designed to automate security testing for AI applications. Aegis provides developers
A paper published by researchers at Carnegie Mellon University's CyLab has introduced a novel jailbreak technique named 'Cognitive Override.' This method bypasses the safety filters of major large lan
The US Department of Commerce's Bureau of Industry and Security (BIS) has enacted a new rule establishing export controls on powerful AI systems. The 'AI Model Export Control Act' restricts the transf
Leading AI firm Cognitive Dynamics disclosed a severe security breach today, confirming that state-sponsored actors exfiltrated core intellectual property, including the weights for its flagship 'Prom
Your 3-minute AI news briefing for Friday, June 5, 2026. The top stories in artificial intelligence, delivered daily.
Direct Preference Optimization (DPO), a key technique for training chatbots, is now being successfully applied to new AI domains like image generation and code synthesis. This breakthrough promises more stable and efficient AI model alignment than ever before.
JetBrains, maker of IntelliJ IDEA, has released Mellum2, a 12-billion parameter Mixture-of-Experts model. This open-source release on Hugging Face challenges existing AI coding assistants. Can it redefine developer workflows?
OpenAI has released a detailed blueprint for US AI governance, proposing a new federal framework to oversee safety, security, and resilience for advanced models. What could this mean for the future of AI development?
The U.S. Department of Homeland Security, in collaboration with CISA, issued a new directive mandating comprehensive security audits and adversarial red teaming for any AI or machine learning system d
The non-profit AI safety organization Aperture Labs has launched LLMGuard, a new open-source Python framework designed to help developers proactively secure their large language model applications. LL