NC State University is committed to being a leader in artificial intelligence (AI). In May 2024, NC State declared to the Board of Governors that NC State is a Data Science and AI University.

This commitment is reflected in the University’s implementation document for its strategic plan, which directs the university to “Consider how NC State will continue to serve as a hub for innovation and application of artificial intelligence (AI) research tools and lead in the implementation of AI in administration, education and other dimensions of running a world-class university.” The AI Advisory Group was charged by university leadership to support and implement this initiative and to provide university-wide perspective and recommendations on the rapidly developing field of AI in education, research and business use.

This page outlines our principles for responsible AI use in research, empowering our researchers to explore the full potential of AI while adhering to the highest ethical standards. It is designed to align with the university’s broader AI vision. Journals, funding agencies, professional societies, publishers and other external entities may have guidance on AI that is more restrictive than this statement. Where applicable, researchers should ethically abide by these external guidelines when publishing papers, submitting grants, managing research contracts or conducting research, even if they are more restrictive than the guidelines described below. Moreover, recognizing that AI is a dynamic and evolving landscape, the principles and specific recommendations in this statement will be revised over time to reflect the most recent understanding.

Principles

Empowerment and Flexibility

NC State champions a culture of experimentation and innovation, where researchers are encouraged to push the boundaries of AI development and applications. Researchers are encouraged to utilize the full spectrum of AI tools and technologies necessary to achieve their research goals. Researchers have the autonomy to select the most appropriate tools for their specific needs. OIT maintains a list of currently recommended AI systems, which is reviewed and updated on an ongoing basis, and instructions on how to purchase AI systems that have not yet received blanket approval. That list is here. Departments and Colleges may have additional guidance to assist in this process. With respect to Open Access AI tools, researchers are encouraged to use their best judgment as allowed by NC State Policies, Regulations and Rules, and informed by this document.

Ethics in Practice

We are committed to fostering a culture of responsible innovation, where ethical considerations are integrated into all stages of research. Ethical responsibility is an ongoing commitment, and researchers should remain vigilant and responsive to emerging risks.

While we encourage cutting-edge AI deployment, we prioritize ethical considerations. Researchers should be mindful of potential biases, privacy and copyright concerns, and societal impacts associated with their AI-driven research.

Although AI models generate novel content, they are trained on vast datasets of human-created content. In some cases, the output may unintentionally contain phrases or ideas that too closely resemble the original source material in the training data, leading to accusations of plagiarism, even if unintentional.

Researchers are expected to stay informed about ethical AI practices and engage with resources and training provided by the university. Because many AI systems are trained on datasets that may include proprietary or copyrighted material, the legal landscape for intellectual property (IP) for AI systems is rapidly evolving.

Researchers should do their best to understand the training data of the AI systems they use and stay informed of emerging IP concerns, relevant laws and all applicable institutional policies. Leaders of teams or groups should consider the effect of AI use on their students and staff, and ensure that the benefits of AI use outweigh the potential risks.

Addressing Bias, Inaccuracy and Accessibility

When applicable, researchers should be mindful of the potential for bias in AI, especially when trained on historical data that already includes biases within the data.
Modern generative AI tools can create content that is not truthful (aka hallucinations); researchers should be aware of the accuracy limitations of the tools they are using, and may wish to implement a human-in-the-loop workflow to validate that AI-produced content is appropriately accurate.

Researchers should strive to ensure that AI outputs and work products are representative of all users, regardless of their backgrounds, and that they foster an accessible environment for the university community by considering the needs or abilities of all individuals.

Responsible Data Stewardship and Privacy

Researchers must handle data used with AI research responsibly, adhering to all applicable privacy laws, university policies, and REG 08.00.03 – Data Management Regulation.

Researchers should be aware of and understand the NC State Data Management Framework, paying particular attention to the sensitivity level of the data they are working with to ensure implementation of appropriate security measures and prevent unauthorized access or disclosure. For research that has specific compliance control plans, use of AI and the specific tools to be used need to be approved before use.

Approved AI tools may have specific agreements with NC State on what type of data is protected and what the company may do with any data or files that are uploaded. Many freely available AI tools provide no guarantee of data privacy.

Respect for Intellectual Property

Researchers must safeguard their own intellectual property (IP) and ensure they do not violate the IP rights of others when using Generative AI. This is particularly important when working with cloud-based, public access AI tools. Inputting unpatented or unpublished information into AI tools, not covered by university enterprise licenses, may lead the University and others to lose their IP rights to that information. Moreover, some AI outputs may build upon or even contain the IP of others. Since the IP status of these outputs is currently an ongoing legal discussion, researchers should proceed cautiously.

A researcher cannot claim IP rights over outputs purely created by AI tools without any involvement of the researcher (per US Copyright Office guidance). AI should not be listed as a co-author, though some agencies and journals may require that the use of AI be documented in detail.

Transparency and Reproducibility

Researchers are encouraged to document their AI methodologies and data sources thoroughly to ensure transparency and reproducibility, and to mitigate the likelihood of fabrication or falsification of data.

Wherever possible, researchers should strive to make their AI models, datasets, and documentation publicly available to promote collaboration and knowledge sharing.
When working with AI in human subjects’ research, the IRB will require that the use of AI and pertinent information related to that use is directly communicated to the research participants in accordance with the NC State IRB’s standards and best practices.

Human Subjects Protection

All research involving human subjects must undergo review and approval by the Institutional Review Board (IRB). This includes any human subjects’ research that involves the use of AI.

To ensure that participants’ rights and welfare are protected, researchers are expected to follow the IRB’s unit standards and guidelines regarding the use of AI in research with human subjects.

Human-Centered Approach

AI should complement, not replace, human expertise and judgment in education and research. Researchers should play an active role and maintain ownership of their intellectual property and professional work, ensuring that technology is used in a way that fosters strong relationships, builds trust and cultivates a supportive environment within the NC State community.

Sustainable AI Practices

The exact environmental impact of AI is hard to quantify since much of the training and development of these tools is proprietary. As such, researchers are encouraged to stay up-to-date on the most recent environmental impact information of their tool use, and to make an informed decision about how to continue with their research.

Moving Forward

This statement will be reviewed and updated regularly to reflect the evolving landscape of AI. We encourage all members of our research community to engage in open dialogue about the responsible use of AI and to contribute to the ongoing development of best practices.

These guidelines are provided and managed by the NC State University Artificial Intelligence Advisory Group.

Resources

From the Office of Research and Innovation

The Office of Research and Innovation supports the research, scholarship and creative activities of the NC State community by promoting collaborative research and an innovative environment.

From NC State University Libraries

NC State University Libraries is here to help you utilize AI in your research. Visit the NC State University Libraries AI for Research resources for more information about how to integrate AI effectively in your research, learn common definitions of AI terms and view available AI tools at NC State.

• This resource serves as a guide for researchers on using artificial intelligence (AI). It defines common AI terminology, lists available tools and computing resources for developing AI models and outlines important risks and ethical considerations.

• This resource provides a guide on AI for researchers, categorizing different AI applications and highlighting computing and data resources available on campus. These resources include a High Performance Computing cluster with GPUs and the OIT’s Research Storage service, both designed to support the intensive computational and data needs of AI projects.

From NC State Extension

NC State Extension provides the groundwork for the strategic, successful use of AI. Review their guiding principles.

NC State University Institutional Review Board (IRB) Guidance

This guidance is intended to assist researchers at NC State University in the participant-centered design and implementation of human subjects’ research, including completion of the IRB application and risks assessment, wherein any artificial intelligence (hereafter, AI) or generative AI (hereafter, GAI) is used. Researchers should utilize this guidance if planning to design and/or implement research with human subjects where AI or GAI is used to interact with participants, used as an intervention with participants, used to manipulate a participant’s environment or used to access, analyze or generate information about one or more participant(s). 

External University Resources

Browse a catalog of AI research resources from other universities.

Ethical Considerations & Guidance

• UNESCO Recommendation on the Ethics of Artificial Intelligence
• IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems
• Partnership on AI
• National Academies of Sciences, Engineering, and Medicine (NASEM)
  • Search for “AI ethics” or related terms.

Research Centers & Initiatives

• National Science Foundation AI Institute for Engaged Learning.
  • The EngageAI Institute researches narrative-centered learning technologies, adaptive collaborative learning and multimodal learning analytics to create deeply engaging collaborative story-based learning experiences.
• Center for AI in Society and Ethics (CASE)
  • CASE aims to facilitate interdisciplinary research — across the humanities, computer science and engineering — along with graduate training experiences. It also aims to promote public and civic engagement by creating humanities-centric collaborations with research institutions and technology companies in North Carolina’s Research Triangle area. 
• N.C. Plant Sciences Initiative
  • The Plant Sciences Initiative brings together the brightest minds in academia, government and industry to drive vital research that increases crop yields, creates new varieties, extends growing seasons, enhances agricultural and environmental sustainability and produces new and improved technology.

More NC State Centers and Institutes

NC State University’s Centers and Institutes provide platforms for the sharpest minds to collaborate, investigate and discover. With more than 40 system-recognized centers and institutes, our faculty are putting their research to work to solve some of the biggest societal challenges while advancing scholarship.