Principles for the Use of Artificial Intelligence (AI) in Education Research
Artificial intelligence (AI) is increasingly shaping the education research process. Researchers are using AI-enabled tools for activities such as coding qualitative data, scoring essays, generating test items, analyzing complex data sets, and drafting research findings. These uses offer opportunities to improve research speed and insights, but they also introduce risks such as responsible data use and transparency, among others.
To help the field navigate these opportunities and risks, AIR convened the AI Education Research Principles Design Lab. Researchers, industry experts, and funders worked together to develop shared, adaptable guidance that complements established research standards and human-subjects protections. AIR then invited public comment from the education research community and revised the principles in response to the feedback received.
Three Principles for Responsible AI Use
The publication introduces three core principles intended to apply across research types, methods, and stages.
| Principle | What It Means |
|---|---|
| 1. Center Human Expertise | Keep human expertise at the core through intentional design, ongoing human oversight, and rigorous human verification. Researchers remain responsible and accountable for decisions, interpretations, and conclusions. |
| 2. Confirm Model Appropriateness | Ensure that model specification, accuracy, and stability are appropriate for the research questions, populations, settings, data, and intended uses. Researchers should evaluate limitations and the consistency of findings across relevant contexts. |
| 3. Operationalize Transparency | Build transparency into the research process by protecting participants, clearly disclosing how AI tools are used and where human judgment applies, and appropriately attributing contributions from models, developers, participants, and other collaborators. |
As AI technologies continue to evolve, the principles will be refined over time.
Moving Forward Together
Realizing the potential of these principles will require collective action across the education research system. Researchers, funders, institutional leaders, institutional review boards, journal editors, professional associations, and practitioners can use the principles to examine and strengthen research standards, practices, and policies.
Read Principle for the Use of Artificial Intelligence in Education Research to learn more.
We welcome additional feedback on the principles. Please share your input through this survey.
