Ron,
I see the point about the condition where joists are only coming in from one side. Unless there is a positive connection between the joists and steel beam, the beam could conceivably buckle away from the joists.
But if the joists are on both sides, each providing a positive compressive restraint against sidesway translation, I would think you could consider the beam braced. The small amount of compressibility wouldn't seem to adversely affect the bracing.
If the beam was not heavily loaded, the lateral force, and stiffness required for lateral stability wouldn't be that significant.
What do you think?
We have typically used wood joists as bracing, but admittedly, we usually have the wood on top of the beam.