May 19, 2025

Webinar Recap: How do we responsibly support researchers with GenAI?

Providing Visible & Invisible Support

This week, we hosted our fifth installment in Atom's series on AI and technology in research administration, featuring the insightful Lori Ann Schultz, Assistant Vice President for Research Administration at Colorado State University. The webinar, "How do we responsibly support researchers with GenAI?", brought together research administrators from across the country to explore the transformative potential of generative AI in the research ecosystem.

The Changing Landscape of Research Administration

We find ourselves at a pivotal moment in research administration. As Lori Ann pointed out, there's an interesting tension emerging in academia: many faculty members express concern about students using AI tools while simultaneously exploring how these same technologies might enhance their own research processes. This presents a unique opportunity for research administrators to guide responsible AI adoption across institutions.

The webinar emphasized that research administrators don't need to wait for institutional licenses to make an impact—there are immediate, responsible ways to leverage available AI tools while safeguarding sensitive information and helping researchers navigate regulatory frameworks.

Key Considerations for Responsible AI Use

Data Governance and Confidentiality

A central theme throughout the discussion was the critical importance of data governance. Lori Ann highlighted several crucial questions research administrators should consider:

• Who controls the data used in AI systems?
• What are the confidentiality expectations for research data?
• How do we protect business-sensitive information when using AI tools?

With AI's ability to collect and process data at unprecedented speeds, maintaining proper data governance becomes even more essential.

Navigating Sponsor and Publisher Policies

The webinar revealed interesting nuances in how different stakeholders approach AI use:

• Sponsors: Many funding agencies have yet to develop specific policies regarding AI use in proposal creation, though restrictions often exist for peer review processes.
• Publishers: Academic publishers typically have clearer guidelines, with some requiring citation of AI use or prohibiting AI-generated content, particularly images.

Understanding these varying policies is crucial for research administrators guiding faculty through the responsible use of AI tools.

Ethics and Bias Awareness

Lori Ann emphasized several ethical considerations that should inform AI adoption:

• Access equity: Ensuring fair access to AI tools across the research community
• Inherent bias: Recognizing that current generative AI models tend to be Western and English-centric, potentially presenting incomplete information
• Environmental impact: Considering the energy usage associated with AI systems as a future consideration

Practical Applications Throughout the Research Lifecycle

The webinar provided concrete examples of how AI can be responsibly integrated into various stages of the research administration lifecycle:

Proposal Development

• Summarizing RFPs and polishing language (using free tools)
• Checking proposals against requirements and assisting with literature reviews (with enterprise licenses)

Opportunity Matching and Award Management

• Matching funding opportunities with faculty based on research activities
• Generating award summary sheets and suggesting language for contract negotiation
• Automating PI acceptance of award terms and AI-driven contract triage

Compliance and Training

• Drafting data management plans and identifying compliance requirements
• Monitoring regulatory changes
• Creating user aids and developing professional development paths for staff

Policy Development and Institutional Support

• Leveraging AI for drafting research administration policies
• Improving prediction and visualization of proposal and award data
• Conducting workload analysis and benchmarking
• Identifying potential faculty collaborations and creating research chatbots

Guiding Faculty on Responsible AI Use

While AI can significantly enhance administrative efficiency, Lori Ann emphasized that research administrators play a crucial role in guiding faculty on responsible use. This includes:

• Creating awareness of institutional data governance policies
• Educating researchers about sponsor and publisher requirements
• Providing safe environments for faculty to explore AI in research development

Q&A Insights

The webinar included a robust Q&A session addressing several practical concerns:

• AI Tool Preferences: Lori Ann shared her experience with Co-pilot and ChatGPT, noting the importance of human expertise in vetting AI outputs due to potential "hallucinations"
• Workload Analysis: AI can help understand staff workloads and plan professional development by tracking proposal types and complexity
• Funding Opportunity Matching: Combining data from proposals, awards, publications, and teaching can improve accuracy in matching faculty with appropriate funding opportunities
• Institutional Guidelines: When using personal AI accounts, it's safest to avoid inputting institutional information and only use publicly available data
• Post-Award Monitoring: Institutional AI tools can help with monitoring grants, particularly for effort and budget deviations
• Environmental Considerations: While still an emerging area of research, early comparisons suggest potential net benefits through efficiency gains
• Increasing AI Adoption: Creating institutional visibility through steering committees and safe environments for experimentation can encourage broader adoption

Atom's AI-Powered Research Administration Platform

Tomer du Sautoy, co-founder of Atom, concluded the webinar by showcasing Atom's platform designed specifically for enhancing research administration. Key features highlighted included:

• Processing RFPs and supporting documents to help researchers find suitable funding
• Generating personalized eligibility scores based on researcher profiles
• Providing an integrated chat experience for extracting information and ideating proposals
• Offering automation features for administrators to create researcher profiles and deliver curated funding opportunities

Moving Forward

As we navigate the evolving research landscape, it's clear that AI tools will play an increasingly important role in supporting both researchers and administrators. By approaching AI adoption thoughtfully—with proper attention to governance, ethics, and responsible use—research institutions can harness these powerful technologies to enhance productivity while maintaining the integrity of the research process.

The growing importance of AI in the research ecosystem presents a valuable opportunity to support growth despite current challenges in the research funding environment. As Lori Ann aptly demonstrated, the key lies in staying informed and providing effective guidance as we collectively explore this new frontier.

For more information about Atom's AI-powered research administration platform, book a demo or contact Tomer at tomer@atomgrants.com.