The Lonely Quest of Scout 752

What if future rescues in space are complicated by the presence of hostile opponants? Messages will be intercepted. Ships can be detected. Sophisticated artificial intellegences are still susceptible to hacking -- inquisition, suggestion, coercion. A human being has no place here.

What is needed is a simple AI, a tiny spaceship, an single-minded mission. Two of those are easy; but how do you build an AI basic enough to get the job done without betraying the mission?