this is really important in the context of debates about AI. The ethos of Open Access is to permit anyone to pass research as data to software.
I don't believe a stance along the lines of: "I don't want my research to be used in AI training" is compatible with making ones research open access.