This essay by Lance Eaton discusses the matter of “academic fracking”, advocating for practices that might be seen as consistent with FAIR-ness:

If we are ever going to get the AI that we hope for—the kind that is insanely helpful, accurate, and insightful, then we are going to need to find a way to transfer all of our captured knowledge over to it. That should ideally be a process that honors and recognizes the countless authors who have something to contribute. It should be celebrated and honored; not hidden away and not mentioned by the companies doing it. 

More: 

https://aiedusimplified.substack.com/p/academic-fracking-when-publishers

Others, as in this post from The Conversation, have also been concerned with extraction, and the recolonizing potential and action, of LLM and unaccountable AI, contrary to the spirit of OCAP(R)  and CARE.