Research NewsletterIssue: ORN-2023-27

The Use of Generative Artificial Intelligence Technologies is Prohibited for the NIH Peer Review Process

NOT-OD-23-149

The purpose of this Notice is to clarify NOT-OD-22-044 on Maintaining Security and Confidentiality in NIH Peer Review: Rules, Responsibilities and Possible Consequences and inform the extramural community that the NIH prohibits NIH scientific peer reviewers from using natural language processors, large language models, or other generative Artificial Intelligence (AI) technologies for analyzing and formulating peer review critiques for grant applications and R&D contract proposals. NIH is revising its Security, Confidentiality, and Non-disclosure Agreements for Peer Reviewers to clarify this prohibition. Reviewers should be aware that uploading or sharing content or original concepts from an NIH grant application, contract proposal, or critique to online generative AI tools violates the NIH peer review confidentiality and integrity requirements. 

Confidentiality and AI Technologies: Maintaining security and confidentiality in the NIH peer review process is essential for safeguarding the exchange of scientific opinions and evaluations. Materials pertaining to an application or proposal, and any other associated privileged information, cannot be disclosed, transmitted, or discussed with another individual through any means, except as authorized by the Designated Federal Officer (DFO) in charge of the review meeting, or other designated NIH official, as stated in the NIH Security, Confidentiality, and Non-disclosure Agreements for Peer Reviewers. The use of generative AI tools to output a peer reviewer critique on a specific grant application or contract proposal requires substantial and detailed information inputs. AI tools have no guarantee of where data are being sent, saved, viewed, or used in the future, and thus NIH is revising its Confidentiality Agreements for Peer Reviewers to clarify that reviewers are prohibited from using AI tools in analyzing and critiquing NIH grant applications and R&D contract proposals. Such actions violate NIH’s peer review confidentiality requirements.

Implementation and Notification: As part of the standard pre-meeting certifications, all NIH Peer Reviewers will be required to sign and submit a modified Security, Confidentiality and Nondisclosure Agreement certifying that they fully understand and will comply with the confidential nature of the review process, including the prohibition on uploading or sharing content or original concepts from an NIH grant application, R&D contract proposal, or critique to online generative AI tools. NIH will also extend this policy to members of NIH National Advisory Councils and Boards and will require such members to certify similar Security, Confidentiality, and Nondisclosure agreements.

Additional Information: Computer technologies that are used for accessibility needs may be granted an exception to this policy. NIH Peer Reviewers must communicate the technology being used with their Designated Federal Officer in charge of the review meeting or other designated NIH official prior to use.

For additional information on applicable laws, regulations, and policies, as well as possible consequences for violations of the NIH peer review rules, see Maintaining Security and Confidentiality in NIH Peer Review: Rules, Responsibilities and Possible Consequences.

Resources: Use of Generative AI in Peer Review FAQs

NSF: National Science Foundation Research Traineeship (NRT); Program Research Experiences for Undergraduates (REU): Sites and Supplements; Assessing and Predicting Technology Outcomes (APTO); Engineering of Biomedical Systems; Cyberinfrastructure Technology Acceleration Pathway (CITAP)

NIH: Cellular and Molecular Biology of Complex Brain Disorders (R21); BRAIN Initiative: Development and Validation of Novel Tools to Probe Cell-Specific and Circuit-Specific Processes in the Brain (R01)

Department of Defense/US Army/DARPA/ONR: FY23 Science & Technology for Advanced Manufacturing Projects (STAMP); Biological Technologies

National Endowment of Humanities: Spotlight on Humanities in Higher Education; Humanities Connections; Climate Smart Humanities Organizations

Can NIH help drive health data access?: The National Institutes of Health released a new data-management and data-sharing policy this year that seeks to address longstanding data interoperability and oversight issues that have plagued the federal government’s response to public health emergencies like the COVID-19 pandemic. Its success will now depend on how effectively the policy can be implemented across the nation's medical research agency and its diverse network of complex IT systems that span hospitals, clinical centers, research institutions and external partners throughout the public health community.  The policy was years in the making, and collaboration across those complex IT systems was a key component from the beginning, according to Cindy Danielson, associate director of systems integration for NIH's Office of Research Reporting and Analysis.  “We are really looking to help enable a cultural shift in research, one in which data sharing is the norm,” Danielson said. “One lesson learned is that community engagement is really key, and NIH has focused on engaging the community while developing the policy, while implementing it and continuing now while it’s in effect.” The agency has since taken a phased approach and recently launched a pilot project with the research community to assess its implementation. The pilot is spearheaded by the Federal Demonstration Partnership, a cooperative of 10 federal agencies and more than 200 institutions focused on reducing administrative burdens associated with research funding. More information is posted on the NextGov website.

________________________

White House unveils fiscal 2025 cybersecurity investment priorities: The White House has released its cybersecurity investment priorities for fiscal 2025, urging federal agencies to adopt key pillars of the National Cybersecurity Strategy in their budget proposals and overall missions. The Office of Management and Budget memorandum issued Tuesday focuses on the same five pillars featured in the national cybersecurity strategy: defending critical infrastructure, disrupting and dismantling threat actors, shaping market forces to drive security, investing in a resilient future and forging international partnerships.  It calls on the federal government to modernize its information technology systems by investing in "durable, long-term solutions that are secure by design" and improving baseline cybersecurity  requirements. The fiscal 2025 priorities also align with the Federal Zero Trust Strategy released last year, which seeks to ensure governmentwide cybersecurity practices are in place and that every access attempt is verified on federal systems and networks. The memorandum instructs agencies to prioritize modernization efforts for systems that are reaching end of life or end of service, when they are typically at their most vulnerable to cyber intrusions. More information is posted on the NextGov website.

________________________

First AI advisory committee report stresses getting regulatory balance right: The National Artificial Intelligence Advisory Committee delivered its first congressionally-mandated report to President Joe Biden that offers recommendations on how to maximize the benefits of implementing artificial intelligence across U.S. sectors and how to manage emerging systems like generative AI. The Year 1 report, dated May 2023 and announced June 22, includes steps prioritizing trustworthy AI systems, new research and development initiatives and more international partnerships focused on aligning standards in AI governance.  NAIAC members, who are appointed by the secretary of Commerce and operate with the support of the National Institute of Standards and Technology, arrived at the conclusion that federal direction will help define whether or not AI technology will have a net positive or negative impact. More information is posted on the NextGov website.

Question:       Can I generate budgets for multiple years from the Year-1 budget in Streamlyne?

Answer:         Yes! You only need to input the Year-1 budget and then click on the “generate all periods” button. Stremalyne will create budget sheets for the remaining periods. You can then go to “summary” under the budget tab to review budget sheets for all periods. You can also change specific budget items that you allocated in Year-1 but you do not want to continue them in the following periods.

More FAQs on Streamlyne: Please visit http://www.njit.edu/research/streamlyne/