Generative AI, Academic Deepfakes, and Epistemic Pollution

Concerns about GenAI’s energy use and physical pollution are well known. This commentary highlights another threat: epistemic pollution—the degradation of our knowledge ecosystem through low-quality, misleading, or fabricated information. Drawing on a recent pseudo-scientific article co-authored by Grok 3, we discuss the rise of academic deepfakes and their contribution to epistemic pollution. We show how GenAI makes academic deepfakes cheaper, more accessible, and harder to detect. The resulting epistemic pollution poses acute risks in fields like Business & Society, which draw on multidisciplinary knowledge and engage with politically sensitive topics. Yet, by extending the analogy between epistemic and physical pollution, we suggest that Business & Society scholars are also well placed to study the problem and respond.