AI Governance

All Data Governance Resources

AI governance is an extension of data governance. While AI systems rely on data, they also generate new data. That generated data is an institutional asset and must be protected just like any other college data.

Examples of AI-related data include:

• Inputs and outputs from AI chatbots such as Vanderbelt’s Amplify, the College’s chat platform, OpenAI’s ChatGPT, Microsoft Copilot, Perplexity, Google Gemini, Anthropic’s Claude, and Grok
• Machine learning (ML) models that use structured and unstructured data for regression, classification, semantic analysis, and clustering
• Training and inference data used in ML models
• Predictions and decisions produced by ML models
• Data shared with or processed by third-party AI systems

Through the College’s data governance framework, we have already established guardrails to protect institutional data and mitigate risk. These include access and sharing controls, human oversight, and training. Many core data governance principles directly overlap with AI governance, including:

• Ethical use
• Transparency
• Accountability and responsibility
• Continuous monitoring and evaluation
• Providing appropriate access based on roles and permissions
• Protection of sensitive data
• Respect for intellectual property

Together, these principles ensure that both data and AI technologies are managed responsibly, securely, and in alignment with the College’s values and obligations.