Institutional AI Governance and Oversight Policy Template
Policy
1. Purpose and Scope
[Institution Name] recognizes that Artificial Intelligence (AI), including generative AI and large language models (LLMs) and similar tools, is rapidly transforming medical education, clinical practice, research, and administrative operations. When utilized responsibly, AI can enhance student learning, improve patient outcomes and reduce administrative burdens. However, inappropriate use can undermine academic integrity, compromise patient privacy, exacerbate health inequities and violate legal obligations.
This policy establishes a unified, institution-wide framework for the ethical, secure and responsible procurement, integration and use of AI. It applies to all students, faculty, clinical staff, researchers and administrators engaged in academic, clinical, scholarly or operational activities on behalf of [Institution Name].
2. Core Guiding Principles
The integration of AI at [Institution Name] must align with our mission of advancing osteopathic philosophy and whole-person healthcare. All AI deployment and usage is governed by the following core and bioethical principles:
Osteopathic & Bioethical Alignment: AI use must be grounded in the principles of autonomy, beneficence, nonmaleficence and justice. AI use must align with the core tenets of osteopathic medicine, where applicable.
Human Accountability (Human-in-the-Loop): AI is a supplementary tool, not a replacement for human reasoning, clinical judgment or empathy. Users remain fully accountable for all work products, academic submissions and clinical decisions, regardless of AI assistance.
Equity and Mitigation of Bias: AI systems must be evaluated to ensure they do not exacerbate existing disparities in healthcare access, treatment or outcomes. Users must actively identify and mitigate algorithmic biases.
Transparency: The use of AI must be transparently disclosed in curriculum development and deployment, academic submissions, research publications and, where appropriate, clinical decision-making.
3. Governance and Oversight Structure
To ensure the safe and mission-aligned implementation of AI, [Institution Name] establishes an interdisciplinary AI Governance Committee.
Separation of Strategy and Governance: While executive leadership will dictate the institution's AI strategy, the AI Governance Committee is strictly responsible for reviewing, assessing and evaluating individual AI tools.
Tool Registration and Assessment: The committee will maintain a centralized inventory of all AI tools used within the institution. Before implementation, tools will be evaluated across multiple domains, including privacy, legal compliance, patient safety, algorithmic bias and IT integration.
Approval Levels: [Institution Name] will maintain a tiered list of vetted AI tools indicating approved use cases (e.g., tools cleared for public data vs. tools cleared for Protected Health Information).
4. Use in Education and Teaching
The primary goal of the medical curriculum is to cultivate competent, compassionate osteopathic physicians. AI must not replace the challenge of proactive effort and self-reflection necessary for true learning.
Faculty Authority: Instructors maintain the autonomy to permit, restrict or entirely prohibit the use of generative AI in their courses.
Syllabus Requirements: All course syllabi must include a clear, specific policy regarding the acceptable use of AI, outlining permitted activities, documentation/citation requirements and prohibited actions.
Academic Integrity: Using AI to generate an assignment and submitting it as one's own original work, without proper authorization and attribution, constitutes plagiarism and a violation of the [Code of Academic Integrity / Student Conduct Code]. Use of AI by faculty and/or students as part of curricular effort must be transparently documented.
5. Use in Clinical Practice and Patient Care
The application of AI in clinical settings carries high stakes for patient safety and institutional liability.
Prohibition of Unapproved Tools: Clinical staff and students are strictly prohibited from entering patient data, Protected Health Information (PHI) or clinical notes into unauthorized, public-facing AI tools (e.g., public ChatGPT). AI may only be used for patient care if the tool has been explicitly vetted by the AI Governance Committee and is supported by a HIPAA-compliant Business Associate Agreement (BAA).
Qualified Human Intervention: Clinical decisions influenced by AI must feature specified intervention points where a licensed, qualified human physician exercises independent clinical judgment to verify the AI's recommendation.
Clinical Documentation: Medical students may not use AI applications to write History & Physicals (H&Ps) or patient notes outside of those specifically embedded and approved within the institution’s Electronic Health Record (EHR) system. Medical students must have training, where required and/or appropriate, for use of AI applications as part of patient care experiences.
6. Use in Research and Scholarship
Ensure responsible use of AI in research by protecting sensitive data, maintaining transparency in authorship and methods, and upholding academic integrity standards.
Data Protection: Unpublished research data, proprietary institutional data and human subject data must never be entered into non-approved or public AI platforms.
Authorship and Peer Review: AI tools cannot be credited as authors on publications, as they cannot take responsibility for the integrity of the work. Furthermore, AI may not be used to evaluate or review confidential materials, such as in the grant or manuscript peer-review process.
Disclosure: Researchers must transparently disclose the use of AI in the methodology or acknowledgments sections of their manuscripts and grant proposals, strictly adhering to the specific guidelines of publishers and funding agencies.
7. Data Privacy and Cybersecurity
The privacy of our students, patients and staff is paramount.
All AI usage must comply with federal laws, including the Family Educational Rights and Privacy Act and the Health Insurance Portability and Accountability Act.
Because public AI models often train on user inputs, [Institution Name] personnel must operate under the assumption that any information entered into a non-enterprise AI tool becomes public domain.
8. Policy Violations
Misuse of AI—including unauthorized data sharing, failure to disclose AI assistance, circumvention of academic requirements or implementation of unvetted clinical tools—will be subject to disciplinary action in accordance with [Institution Name]'s Academic Misconduct Policy, HR policies and Professionalism standards.
9. Policy Review
Because AI technology is rapidly evolving, the AI Governance Committee will review and update this policy [Annually / Bi-Annually] to account for new capabilities, emerging legal regulations and shifts in the healthcare landscape.
Sample Email for Rolling Out AI Policy
Subject: Announcing [Institution Name]’s New Institutional AI Governance and Oversight Policy
Dear [Institution Name] Community,
As we navigate an era of unprecedented technological advancement, Artificial Intelligence (AI) and generative AI tools are rapidly transforming the landscape of medical education, clinical practice and research. At [Institution Name], we believe these tools hold incredible potential to enhance student learning, improve patient outcomes and streamline our daily operations.
However, we also recognize the profound ethical, legal and privacy challenges this technology presents. To ensure we are integrating AI in a manner that is safe, secure and firmly rooted in our osteopathic mission and core bioethical principles, we are proud to announce the release of the [Institution Name] Institutional AI Governance and Oversight Policy.
This policy applies to all students, faculty, clinical staff, researchers and administrators. We ask that you review the full policy here: [Link to Policy].
Key Highlights of the New Policy:
Human Accountability: AI is a powerful supplementary tool, but it will never replace human reasoning, empathy or independent clinical judgment. Users remain fully responsible for all work products, academic submissions and clinical decisions.
AI Governance Committee: We have established an interdisciplinary committee to review and vet AI tools before they are implemented on campus or in our clinics. The committee will maintain a tiered list of approved tools based on data security levels.
Education & Academic Integrity: Faculty maintain the authority to permit or restrict AI use in their courses, and all course syllabi must now include a clear AI usage statement. Submitting AI-generated work without proper authorization and citation constitutes a violation of our Code of Academic Integrity.
Patient Care & Data Privacy: Entering patient data, Protected Health Information or student education records into unauthorized, public-facing AI tools (like the public version of ChatGPT) is strictly prohibited. AI may only be used in clinical settings if the tool has been explicitly vetted by our AI Governance Committee.
Research Transparency: Researchers must ensure unpublished data is kept out of public AI platforms, and the use of AI must be transparently disclosed in all academic publications and grant proposals.
Next Steps & Support
We understand that adapting to these new guidelines will be an ongoing process. In the coming weeks, the AI Governance Committee will be hosting [Town Halls / Q&A Sessions / Training Workshops] to discuss the policy in greater detail, demonstrate approved tools and answer your questions.
If you have an AI tool you would like to use for university business, research or clinical care, please submit it for review via [Link to IT/Governance Committee Intake Form]. For immediate questions regarding the policy, please contact [Email Address/Contact Person].
Thank you for your continued dedication to excellence, innovation and whole-person healthcare. Together, we can harness the power of AI to advance our mission while safeguarding the trust of our students, patients and community.
Sincerely,
[Name/Title of President, Provost, or Chief Medical Officer] [Institution Name]