It might have been a mix of several factors causing the pitch up.
As i´m unable to think in bits and bytes i´ll do the traditional approach (aerodynamics and human behaviour).
My first thing to do after takeover from autopilot or student (i have the aircraft) was o get a feel for the flightcontrols. Grip the stick and do some small inputs to see if it acts correctly and autopilot is deactivated or the other guy has his hands of the stick. That was an automated habit.
Flightcontrol response is different depending on speed, altitude, CG, and in case of FBW on LAW. As somebody stated before the flightcontrol response of the A330 is more sensitive in ALt2 than in normal, and CG in cruise is more aft then during TO and LDG, where stick time is more present.
AF447 was not flying in undisturbed air, also there is no proved evidence yet that it was turbulence in a significant severity. However, the possibility that there was turbulence affecting smooth flight can be taken as assured.
Hand flying in FL350 seems to be a seldom live expierience by an FO, if it gets necessary in an event the CPT probably will do the flying and then the problem will be quickly solved or the flightlevel will be reduced to get better lifties. So it can be assumed, that whoever was flying at 02:10 could not rely on a lot of handflying expierience in Alt LAW in FL350, and not at all in less than smooth air.
Let me get to my point.
2:10:05 - 2:10:20
AP and AThr disconnect and Alt Law, and a PF / FO with minimal handflying expierience gets the AC in his lap with a roll to the right, takes the stick and tries to get a feel for it like in lower altitude, intends to counter the roll and to set the pitch 5° NU for the UAS procedure, simultaneously trying to counteract some mild or not so mild turbulence. He gets it wrong, his input is to big aggrevated by some outside lifties, and the pitch gets higher then intended. AOA being 4° at the beginning, raised quickly to above 10, triggering the stall warning (yes, i think that was a valid stall warning). Same time he is distracted by the cacophonie of the warning systems while trying to figure out what´s happening.
Only 15 sec passed, we are now at FL375 and Pitch is 10°, AOA not known but somewhere around 10°.
Imho it does not need any technical glitch to get in this high pitch there, not enough expierience for such a situation, some turbulence and a bad day will be sufficient.
02:10:20 - 02:10:45
Now the nose is up and PF tries to get it down again, but PF is over cautious from unintentional NU expierience and his ND input is not big enough to get the nose down quickly, but sufficient to keep the aircraft in an somewhat less then 1G slight upward trajectory. There is again speed indication at 215 Kts, but is it valid? The aircraft looks stable, time to sort out the warnings, ecam.....
Only 20 sec have passed, speed is down to 200 and the climb stopped around FL 375, balistic trajectory basically ended and AOA increased and trigggerd the stall warning in the end. AOA at the beginning before unloading somewhere around 10, then down to 4 after unloading, increasing above 6 again due to decaying airspeed and stopping of climbrate (change of relative wind).
Everything is gone haywire within 45 seconds from routine to nightmare (actually we dont see it that way from our armchair, but the PF might have seen it that way). AF447 is now off track off heading turning to the bad zone of the WX area, 2.500 feet above cleared altitude without contact to ATC, at low speed and all good helpers for cruise flight are on leave when mostly needed. Aditionally the most expierienced help, the captain unfortunately is not on the flightdeck.
02:10:51 - 02:11:40
Stall warning sounds again and the indicateing valid speeds are decreasing through 200kts, and (BEA does not state it, but it looks logical to me, therefore i take that guess) the altitude starts to decrease despite the fact, that pitch is still 10° NU. Natural instinct says power (TOGA) and keep the ship from falling into the bad WX area below (SS up), also we know, that this was the big mistake. The aditional thrust kept the altitude from decreasing initally and brought the ship even to FL380 (again overcommanding of SS and THS trimming), but now gravity strikes again, even the pitch increased from 10 to 16°, the aircraft starts its fall in this attitude (stall) out of FL380 down through FL350 at 02:11:45, at what time the captain entered the cockpit.
I stopp there, because the question was why pitchup. If there are computer glitches aiding this pitchup, we will know later. I for myself stated my oppinion how the events happened above, using
A33Zab´s very helpfull chronology of events.