Artificial intelligence, epistemic authority, and emerging risks in veterinary clinical decision-making

Area of Science:

• Veterinary Medicine
• Artificial Intelligence
• Clinical Decision-Making

Background:

• Artificial intelligence (AI) is increasingly integrated into veterinary medicine for diagnostics and decision support.
• Veterinary clinical decisions are influenced by animal welfare, owner preferences, economics, and legal factors.
• AI's impact extends to the structure of clinical reasoning, necessitating evaluation beyond mere performance.

Purpose of the Study:

• To theoretically review the potential risks of AI in veterinary clinical reasoning.
• To examine AI's influence on epistemic authority and professional judgment.
• To propose a framework for critically evaluating AI in veterinary practice.

Main Methods:

• Theoretical narrative review based on literature searches (PubMed, Scopus, Google Scholar).
• Conceptual examination of clinical decision-making, explainability, automation bias, epistemic authority, and veterinary ethics.
• Identification of potential risks requiring empirical testing in veterinary clinical settings.

Main Results:

• Two primary risks identified: authority delegation and epistemic-normative reshaping.
• Authority delegation: AI outputs may become de facto references, narrowing independent clinical judgment.
• Epistemic-normative reshaping: AI may alter the informational and ethical framework of decisions, especially in complex cases like euthanasia or off-label use.

Conclusions:

• AI in veterinary medicine presents risks beyond accuracy, including potential shifts in clinical reasoning.
• AI should be viewed as a critically reviewable support tool, not an epistemic authority.
• Further empirical research is needed to understand and mitigate these risks in practice.