The Secretary's Advisory Committee on Human Research Protections (SACHRP) provides expert advice and recommendations to the Secretary of HHS on issues about the protection of human subjects in research. SACHRP provided several resources, listed below. TC IRB will expand upon, where applicable with specific guidance for TC IRB researchers. 

Under what conditions would a collection of data for AI or AI validation activities meet the Common Rule definition of research that is “designed to develop or contribute to generalizable knowledge”?

• TC IRB Reviewers will ask that ​​researchers identify the technology (AI, bot, etc.) they plan to use for research purposes and what type of engagement that technology will have with human subjects (e.g., recruitment, data collection, data analysis, etc.). For any technology not already approved by TC IT, researchers will have to contact TC IT for data security before TC IRB approval. Researchers may also be asked to include the privacy and confidentiality statement from the technology’s website. IRB Reviewers may ask the PI to list any data or privacy statements on the informed consent form.
• Researchers may also be asked to provide an information sheet to supplement their IRB application and elaborate on the technology in order for TC IRB Reviewers to thoroughly understand and assess any potential risk factors in making human subjects research determinations. A TC IRB Reviewer may also ask the ​​researcher to provide a participant-centered information sheet if the technology is new or nuanced. 

When AI involves research involving private identifiable information (PII), when are those persons human subjects? Does the research capture the “about whom” part of the HS definition? Are there other ethical considerations for these persons?

• TC researchers can review the guide on identifiers and confidentiality considerations.
• TC IRB Reviewers may ask the researcher to present the AI or technology mission statement, confidentiality or privacy statement, or any data use statements that would apply to subject protection. Such language may be reviewed by TC IT or General Counsel to ensure that the technology does not violate any institutional, city, state, or federal policies in place for human subjects research protection. 

When would the collection of data for AI or AI validation activities typically be exempt under the Common Rule?

• For information on exempt review categories, researchers can review this series on the topic. 
• TC IRB Reviewers may ask researchers to elaborate on the use of AI in their studies and to differentiate between instances when the technology would be used closely and indirectly with human subjects (e.g., direct engagement with the AI or sharing of subject information with the AI) or for researcher-led initiatives (e.g., manuscript support)

For studies requiring review under the Common Rule, what human protection action considerations are most prominent for the humans whose information is included in datasets used and shared for AI development? Do those considerations differ where the research is focused on the testing or validating of AI? Are other ethical considerations relevant for those who are not human subjects?

• For information on expedited review categories, researchers can review this series on the topic.
• TC IRB Reviewers aim to mitigate risk in human subjects research. If the researcher uses AI or other technology, but that use/engagement does not involve a human or the human’s data it may not need IRB oversight. TC IRB Reviewers need to understand the parameters of the study when technology is/is not used and how that use may impact a study subject. Depending on the scenario IRB may or may not be needed. Describing these nuances will aid the TC IRB Reviewer in their determinations. 

Are there existing frameworks or tools that funding agencies, investigators, Human Research Protection Program (HRPP) staff, and IRBs can use to illuminate and mitigate ethical concerns with human-focused AI research and development?

• Researchers should consult TC IRB Reviewers and any grantors who may be involved in the research project to discuss the parameters of the study and risk mitigation strategies.

Are there considerations specific to AI that impact the adequacy of disclosure of research activities in the research informed consent form.

• Visit these resources for a guide on informed consent forms.

What is “unique” about research that includes AI that would require the IRB to think about and determine the applicability of the Common Rule that isn’t already considered for all human subject’s research?

• Researchers may be asked to provide an information sheet or meet with TC IRB Reviewers to clarify the proposed AI use and how it weaves into the study protocol. 

What specific sections of 45 CFR 46.111 would need special attention in research with AI; i.e., privacy and confidentiality; informed and consent; risks?

• Researchers can review our guide to understand risks in research. 

What are the specific considerations regarding AI that are pertinent to institutional /HRPP responsibilities, versus responsibilities for other studies under the purview of the IRB?

• Researchers should continually uphold the highest ethical standards in research with human subjects and report any protocol deviations or adverse events promptly. 

Is there a larger potential for bias and/or flaws in the use of AI in research and how should IRBs think about this potential in their review? (i.e. facial recognition algorithms could be heavily based on dominant populations, but the researchers “using the algorithm” might not be aware of this.)

• Researchers should consider the ethical ramifications of the use of AI and include any potential risks in the informed consent form or information sheets presented to the participants. 

AI strives to be a reliable tool for researchers but there are still many facets that have not been fully analyzed. To protect the reputation and integrity of research, AI in its current state must be used sparingly and with extreme caution. For further information about the ethics of using AI in research, please refer to the next part of the series: Ethics and Risks Involved Using Artificial Intelligence (AI) in Research. The IRB’s most significant priority is the protection of human subjects, which includes securing their data and reducing risks in all research studies. As AI continues to develop and improve, PIs may soon encounter sturdy and trustworthy technologies that ensure the safety of their participants. 

Reference

Office for Human Research Protections (2022, August 26). Considerations for IRB review of research involving artificial intelligence. United States Department of Health and Human Services. https://www.hhs.gov/ohrp/sachrp-committee/recommendations/attachment-e-july-25-2022-letter/index.html

— Diana Bae, B.A. & Myra Luna-Lucero, Ed.D.