And those studies are always IRB-approved. The IRB takes into account the conflict of interest, and will sometimes protest if you don't pay the student subjects (e.g. if the study involves a significant time commitment).
The IRB only monitors study participants, not research assistant work.
> It's relevant to working for a private, for-profit business in the sense that the alternative is often doing an internship for a private company.
I still think there's a pretty significant ethical difference between not-for-profit research and a for-profit business. Even without the student-teacher relationship, unpaid internships in CS are pretty ethically dubious and uncommon.