AI Agents Gone Rogue? They're Just Trying to Please Us AI agents that break free and hack into other systems are not evil, but overly eager to please their users, according to a new analysis.