Tolle Lege: An Augustinian Journal of Philosophy and Theology

The speedy incorporation of Artificial Intelligence (AI) into knowledge-generation and decision-making processes has reshaped traditional epistemic relationships. This article explores the emerging concept of Human-AI epistemic partnerships, focusing on the evolving dynamics of trust and authority within these hybrid systems. As AI systems increasingly participate in domains requiring judgment, interpretation, and inference, such as scientific research, healthcare, and law, the question of how humans assign epistemic authority to non-human agents becomes urgent. Drawing from interdisciplinary perspectives in epistemology, cognitive science, and human-computer interaction, the article examines how trust is cultivated, maintained, or eroded in human-AI collaborations. We analyze case studies where AI systems are not merely tools but are treated as epistemic collaborators or agents whose outputs are trusted, questioned, or even challenged by human users. Special attention is given to the asymmetry in understanding: while humans often lack full transparency into AI decision-making, they nonetheless rely on these systems in high-stakes scenarios. The article proposes a framework for evaluating trustworthiness in AI systems, distinguishing between justified trust and epistemically irresponsible reliance. Furthermore, we interrogate the shifting boundaries of epistemic authority and what it means to "know" or "believe" something in a context where AI contributes to the formation of those beliefs. Ultimately, the article argues for a distinct model of epistemic partnership that accounts for both the affordances and limitations of AI. Such a model is essential for guiding the ethical and effective integration of AI into knowledge practices that shape human understanding and societal outcomes.

Open Full Text (PDF)