The motors push the handles out from the flush position. What you are hearing is the motor return to the flush position.
What people here are saying is that the handle is attached to the motor by springs.
If the motor was pulling the handle into the door, you would not be able to stop it without damaging the motor.
One way you could test this is to hold the handle to prevent it from retracting. The motor sound continues even while the handle doesn't retract. Likewise, when you do let go, the motor doesn't start up again, the handle simply snaps into the flush position.
Ah, got it; thanks for the clear explanation. I can confirm that my handles do behave the way you've described. If I forcibly hold the handle in the fully extended position (without pulling on it further, which would open the door) and then press the fob to retract the handle, the motor operates normally and the other 3 handles retract to flush, while the handle I'm holding stays out where I'm holding it. Once the motor stops, letting go causes the handle to snap back in quickly, and with more apparent force than when I just allowed it to close on my fingertip gradually.
One hypothesis is that my handle springs are under greater tension than (most of) yours, or out of tolerance in some way. Another hypothesis is that our handle springs have identical tension, and the difference is strictly due to pain tolerance. In either case, the force gauge should be able to quantify this and provide some numbers for comparison. (And yes, I do intend to ask the SC whether the handle tension is within spec when I have my next appointment.)
I note that only one finger (my middle finger, which is the longest and also the thickest) was actually caught and pinched. All the force was applied to it at a single point. If I had been gripping the handle more securely with four fingers, the force would likely have been distributed across my hand.