Verifiable AI for Academic Research

Abstract

This seminar explores how to make AI-assisted academic research verifiable, meaning inspectable, reproducible, and defensible by turning vague prompting into bounded, auditable workflows. It offers practical methods for building audit trails, testing workflow reliability, and documenting AI use in ways that support publications and scholarly integrity.