Anthropic Warns Small AI Models Could Self-Replicate
Anthropic has published research suggesting small AI models could soon self-replicate on consumer IoT devices without needing a datacentre. The warning points to a new class of security risk that could affect smart home devices, small businesses and defenders alike.
Mistral’s ‘Le Chonk’ model signals Europe’s AI ambitions
Mistral has unveiled a new flagship AI system nicknamed Le Chonk. The company says the model is intended to compete with leading open-weight systems and strengthen Europe’s place in frontier AI.
A Beginner’s Guide to Everyday Cybersecurity Defence
Cybersecurity can feel technical, but the foundations are practical. Learn the everyday habits and core concepts that help you spot risks, protect data and work more confidently online.
Google limits Gemini 4 Argon access over safety concerns
Google has launched its latest Gemini model, Gemini 4 Argon, but is not opening it to the public straight away. The company says a phased release is needed while it strengthens safeguards against misuse.