Summary

The NIH Office of Data Science Strategy (ODSS) sought to develop practical and trusted guidelines for the ethical use of AI/ML in biomedical and behavioral research. With emerging technologies presenting complex ethical challenges, ODSS required a process that integrated expert insight with broad community engagement. BioTeam partnered with NIH to design and deliver a strategic workshop that convened thought leaders across government, academia, and industry. The result was a community-informed framework that will guide the ethical conduct of AI/ML research and support NIH’s leadership role in shaping responsible data science practices.

Services Provided:

Strategic planning from initial concept through execution

Design, coordination, and facilitation of a large-scale hybrid workshop

Development of background materials, tools, and agendas for participant engagement

Post-event synthesis and publication of a comprehensive workshop report and guidance

Challenge

ODSS recognized the urgency of establishing ethical guardrails for AI/ML in biomedical research before practices became entrenched without oversight. To succeed, the initiative needed to:

Engage a broad and diverse community of experts and stakeholders

Build consensus around emerging ethical issues and practical solutions

Translate dialogue into actionable guidance with clear ownership and credibility

The challenge lay in designing an event that was not only informative but also participatory, capturing the collective voice of the biomedical research community while maintaining focus on key strategic outcomes.

Approach

BioTeam collaborated closely with ODSS to design an interactive workshop, Toward an Ethical Framework for Artificial Intelligence in Biomedical and Behavioral Research AI/ML. Our team:

Researched and prepared background materials to frame the discussion

Identified and engaged leading voices from AI/ML, bioethics, medicine, and data science

Designed a detailed agenda with plenary presentations, breakout sessions, and interactive tools (SWOT analyses, polling, idea mapping) to maximize participant contributions

Supported hybrid participation by integrating both in-person and virtual attendees

Captured and synthesized insights into structured outputs for NIH’s strategic planning process

Outcomes

The workshop convened experts across ethics, computer science, clinical research, and data management to address pressing issues in AI/ML. Discussions focused on five priority areas:

Data Sharing for General Reuse

Foundation Models

Multimodal Data

Synthetic Data

Proxy Variables

Participants identified key challenges and proposed approaches for each area and we consolidated them into a workshop summary. The summary was used to inform draft guidance documents. This collaborative process established a foundation for NIH to develop and share community-driven guidelines that reflect both technical considerations and ethical standards in AI/ML research.