Steven Sinofsky on the middle ground nobody wants: AI stays legal, and every agent action comes with a safety warning. "My biggest fear is that Europe decides that GDPR was the best thing ever, and they're going to just GDPR AI." "AI will be fine. It'll have one prompt for when text gets emitted that says, 'This vendor is emitting text, and it's probably wrong. Yes or no?' At least in North America, you're not going to censor speech. In Europe, they still will. They'll have filters and…