The reasons these journals are called predatory is because they prey on unwary researchers looking for places to publish their work, tricking them into publishing with them rather than in legitimate places. So just because a paper is in a predatory journal does not mean that it is necessarily bad work.
However, this paper still needs to be treated with greatly heightened suspicion: it should be assumed that it has not had any peer review, and that the authors are unfamiliar enough with the area and with the normal practices of science that they could be tricked by a predatory journal. Alternately, they may have sent the work to the journal intentionally because they needed another line on their CV and knew the paper was not going to survive real peer review.
Thus, although such a paper might be legitimate, it is also very likely to be of low quality and may well even be fraudulent or simply nonsense. It can also cause trouble for you if you cite it, as readers may think that you are not informed enough to tell the difference between real research and predatory crap.
I would thus suggest treating the paper like a random PDF found on the web: it might be usable as a primary source (e.g., like a newspaper opinion piece or a personal essay), or to point you for looking for similar information elsewhere. If this information has only ever been published in this one paper in a fake journal, however, it should be considered to effectively not be published at all: the context in which it has been found casts so much doubt on its likelihood of being legitimate, and anything substantive will likely develop multiple real publications over time.