Can prosecutors prove intent when AI goes rogue? The Information’s @leomschwartz relays Cognition CLO Paul Grewal’s view: “It would be very difficult for them to prove intent just because obviously the agents were acting in a rogue manner.”