Through the Looking-Glass: Transparency Implications and Challenges in Enterprise AI Knowledge Systems