The Need for Prospective Integrity Standards for the Use of Generative AI in Research

Area of Science:

• Scientific research regulation
• Medical research oversight
• Emerging technology in science

Background:

• Federal government historically balances research support with public protection.
• Regulatory frameworks differ across clinical care, human subjects research, and research integrity.
• Existing regulatory models face new challenges from advanced technologies.

Purpose of the Study:

• To address the regulatory challenges posed by Generative Artificial Intelligence (GenAI) in scientific research.
• To propose a shift in regulatory approach for GenAI in research.
• To ensure research integrity and public safety in the era of AI.

Main Methods:

• Analysis of current regulatory frameworks for scientific and medical research.
• Identification of challenges presented by Generative Artificial Intelligence (GenAI) tools.
• Argument for a revised regulatory strategy based on community standards.

Main Results:

• Generative Artificial Intelligence (GenAI) tools introduce risks of increased fabrication, falsification, and plagiarism.
• Retrospectively identifying GenAI-induced errors and intent is challenging.
• Current disparate regulatory models are ill-equipped for GenAI's unique challenges.

Conclusions:

• A move towards prospective, community-defined standards for GenAI use in research is recommended.
• Establishing clear guidelines proactively can better manage risks associated with GenAI.
• This approach aims to maintain research integrity and public trust.