Correct, A GFCI has an extra current coil that injects a 120hz signal into the circuit to cause a trip as soon as power is supplied to the GFCI if the neutral is grounded or makes a connection to another neutral on the load side, it also causes a trip if the hot makes connection to another hot of the same phase, this was added to GFCI's sometime in the late 80's or early 90's, before this if a GFCI had a grounded neutral or it was connected to another neutral on the load side you would only get the trip after a load was turned on and cause at least .005 amps to flow on the other path, our testers would not cause this to happen which made trouble shooting GFCI's back then much harder, now AFCI's that also Incorporated a GFP (30ma) also have this problem and will not trip on a grounded neutral unless enough current is placed upon the the circuit to produce an imbalance of .030 amps which is even higher then the older GFCI's.
So what I'm trying to say is we expect a AFCI to respond the same as a GFCI when the two are not the same, just because the AFCI didn't trip with a grounded neutral doesn't mean the AFCI was defective, it could mean that not enough current was placed upon the circuit that produced enough current to cause a imbalance of .030amps
As stated above, a grounded neutral will not always cause a trip on a AFCI, I have had many small electronic appliances and wall worts not provide enough current to cause a AFCI to trip, even my wiggy when applied across the line and neutral will not produce enough current on the neutral to cause an imbalance of 30ma, so any time you suspect a grounded load neutral or one that is connected to another neutral path back to the source then apply a load of at least a couple amps and it should trip.
Below is a PDF on how most modern GFCI's operate, it tells about the function of the 120hz injection coil that the AFCI does not have.