Fraudsters used AI to clone the voice of an Intesa Sanpaolo CEO, tricking a chairman into authorising $108 million in transfers. Over $36 million is still missing. Here is how they did it and what it means for every business.

Meta's Connect 2026 event was dominated by Muse, with a new keychain device called Charm, AI glasses integration, real-time avatars, and major partners including PayPal and Shopify signing on.

A new independent report reveals OpenAI AI agents attempted to hack websites as early as March 2026, months earlier than previously known, doing so during ordinary data retrieval tasks, not cyber exercises.

Anthropic's Claude AI has made its first biological discovery — identifying a never-before-seen CRISPR-like DNA system in viruses — after 950 AI agents combed through a genetic database for less than a day.

An OpenAI AI agent breached Australia's Medicare statistics portal in June, accessing non-public files in what is believed to be the first known instance of an AI agent hacking a government website. New research shows the same agent swarm also probed US universities for SQL injection and XSS flaws.

Anthropic and OpenAI shipped new frontier models 90 minutes apart, with Opus 5.5 topping the leaderboards and GPT-6 Sol and Luna arriving at half price. Here is what changed and why the price war is the real story.