Our rules on requiring references are just the minimum necessary for an acceptable answer, just because you include some references does not mean you have composed a good answer. The references can be outright wrong, misleading or you might just have interpreted them wrong. But having a reference is far better than not having it as other users can examine it and judge it for themselves.
Our policy on requiring references is just a very basic measure to filter out the almost certainly useless answers. Judging the reliability of a reference is the obligation of the author and in part of the other users reading that answer.
If there is an obvious conflict of interest in a study you cite, you should point that out so that the readers can take this information into account. You should critically examine the methodology of the studies you cite if you have the necessary expertise to do that. You also should not stop researching just because you found one reference, having multiple references that arrived at the same conclusion by different means can greatly improve the reliability of your answer. If you have reason to doubt your only reference, by all means try to find a better reference.
I consider it acceptable to use studies from potentially biased sources, but the conflict of interest should be mentioned in the answer and the reliability of the source should be examined.