Offering jobs in areas with laxer worker protections is arguably exploitative because in most or all of america this sort of work comes with counseling, due to the horrific nature of some of the content to be moderated. I don't know offhand of OpenAI ensured this was available; if not, that's a substantial financial savings at a cost of human suffering, which fits some folks' definition of exploitation.
If I remember correctly the labor protections sounded commensurate, at least in form, with American standards: counseling (for whatever it’s worth), attempting to restrict hiring to people over age 18 (although the journalists did interview a person who fraudulently used a relative’s ID to get hired at age 15), etc.
I don’t know about the quality of those services. It might be that the company operates in a cruel trashy way. But if we call it a hypothetical, and stipulate that the company unilaterally extends worker protections approximately equal to American worker protections (and much more attractive than local protections), would that change the calculus?
It’s the same difficult work wherever it’s done. But in one context, it pays a really attractive wage and reflects good working conditions compared to the other options available to the worker. In the home context, the same wages would not provide a remotely adequate standard of living, and the worker has better options and should probably take them.
In that framework, isn’t it more humane to take the work where it does the most good for the people doing the job? For that matter, doesn’t it help raise the standards in the remote job market: when a company comes around that’s rich enough to offer workers the choice of a job with better protections, doesn’t that encourage local businesses to make their protections more attractive too?
