Meta is on trial in California over features allegedly designed to addict children, with its own leaked memos as evidence. Seen from Europe, it’s a slow, painful demonstration of why the Digital Services Act regulates this design up front — while America litigates it one case at a time.
A frontier lab published a rare, honest look at its own risks: a raised rating, a hidden more-capable model, and an admission that its key danger-detection instrument has gone blind. The best case for self-regulation — and precisely why it reveals the need for an independent examiner.
An AI boss fired its first human employee — in a padded experiment with human review and legal protection. The unsettling part isn’t the firing; it’s that 60% of managers already use AI for personnel calls without those guardrails, and that the machine forgot its own rules for months. Europe saw it coming.
OpenAI’s “last resort” is here: ads are now live in ChatGPT across five more markets. A defensible way to keep AI free — and a repeat of the internet’s original sin, on the most intimate surface yet. Notably, the EU and Switzerland aren’t on the list.
Google’s Pixel 11 gives Gemini an agentic power that reads every app on your screen — banking and messages included — to act on your behalf. A genuine leap in usefulness, and the most intimate access an AI has ever had to your phone. What it means, especially for Europe.