Yes, the vacuum chamber is doing work.
Since the principles should be the same whatever the air pressure, let’s consider the case where the air pressure is so low that there is just one air molecule.
The molecule happens to be moving in a direction which takes it into the chamber. At this point nothing has happened. No force has been exerted by anything on anything. No work has been done.
The molecule reaches the far wall of the chamber and bounces off it. Thus its momentum has changed from $-mv$ to $+mv$ - the wall has done work on it - and accordingly the momentum of the chamber itself has changed by $-2mv$.
Some time later, the bouncing particle reaches the near wall off the chamber and bounces off it. This puts everything back to how it was at the beginning. The work done by the chamber reverts to $0$.
This reflection process can go on indefinitely. I have of course oversimplified, assuming a 1-dimensional chamber; but scaling up to 3 dimensions with possibly oblique reflections does not change the principle.
Now you can scale this up to more air molecules. At any given time, half of them are in the “having been reflected an odd number of times” state (thus, having had work done on them) and half are in the “reflected an even number of times” state, thus having had no work done on them.
In total, therefore, the incoming air has had work done on it by the vacuum chamber.
It is possible to approach this question by considering the momentum of the flowing air rather than individual molecules, and that is probably easier for quantitative calculation. But the fundamental principle is made more obvious by considering the molecular level.