AI Usage Cards: Responsibly Reporting AI-generated Content

The paper proposes a three-dimensional framework for responsible AI use in research consisting of transparency, integrity, and accountability — these dimensions help researchers and institutions evaluate whether AI tools are being used ethically rather than simply adopting them without consideration.

"AI Usage Cards" provide a standardized reporting format for disclosing how AI was used in scientific research, similar to existing disclosure practices in other fields, allowing the research community to compare different approaches and establish shared norms around responsible AI adoption.

Current guidelines for responsible AI use in specific domains like robotics or medicine don't transfer well to scientific research, highlighting a gap that this work addresses by creating a framework specifically designed for the research context where AI-generated content can be indistinguishable from human work.

The paper acknowledges that while we need more time to fully understand AI's benefits and harms, rapid and widespread adoption is already happening — the proposed cards and framework offer a practical interim solution to at least document and reflect on how AI is being used rather than leaving it completely untracked.

The free questionnaire-based tool for generating AI Usage Cards in machine-readable formats aims to make responsible reporting accessible and comparable across research fields, potentially enabling researchers to learn from each other's practices and identify emerging best practices.