According to this answer to an earlier question of mine, the positions of the angle-of-attack sensor vanes on the A330/A340 are dependent not only on AoA, but also on airspeed (and, according to @ymb1's comment, this is true for all aircraft, not just the A330/A340):
1. AoA vanes are speed sensitive
An AoA vane is in essence a weather vane, but due to various aerodynamic interactions it's not perfect. From flight testing they would have worked out how speed affects the position, and based on that the sensed position is corrected by speed input.
This doesn't make sense - the airflow over the wing is qualitatively the same for a given angle of attack, no matter what the airspeed (as long as you stay below the first critical mach number, at which point you get shockwave formation and all sorts of other goodness), so how can the airspeed affect the readings from the AoA sensors? Shouldn't the airspeed merely affect how long it takes for them to weathervane in the direction of the local airflow?