The article discusses the introduction of PEEL, a framework aimed at enhancing epistemic accountability in AI research by integrating traditional analysis tools with large language models. It highlights the need for careful design in AI tools to address inherent distortions and maintain research integrity.