I'd say that it's the transfer case shift relay or something in the GEM calling for the shift at the wrong time. The motor itself has no power to it statically, so it's not going to spontaneously move by itself. It must be receiving power to make the low-high or high-low shift, and that power is usually fed from the fuse, through the relay, and the trigger for the shift comes when the GEM grounds the relay.
Soooo.... I see a few different possibilities:
-The first is a bad GEM incorrectly calling for a shift.
-The second is a short to ground in the wiring that triggers the high-->low shift. Those two wires are in the BR/PK and GY wires that run between the transfer case shift relay and the GEM. I'm not sure which of the two triggers the high-->low shift offhand.... somebody should be able to tell us.
-Third is (and this is a VERY remote possibility) is a failed transfer case shift relay, AND a bad diode in the relay. That *could* potentially allow a back-feed of power from the Torque-On-Demand relay through the infamous brown wire to the transfer case shift relay, through the shift motor. The odds of both parts of that relay failing are VERY remote IMHO.
As a temporary fix, you can remove the motor, turn the shift shaft back to the high position, re-set the motor by applying power to pins #14 and 15 in the plug on the motor (orange and yellow wires) until the motor is in the proper position, and then reinstalling the motor, but leave it un-plugged. With the connector unplugged, the motor can't move. If you still experience the problem, then it's a mechanical problem with the transfer case, for sure, but I doubt that's the case here.
Let us know what you find.
-Joe