Trust Layers for AI in Enterprise Systems

This technical report presents a framework for integrating trust layers into generative AI systems within enterprise environments. It addresses the challenges associated with connecting AI agents, such as ChatGPT, to internal data sources like SharePoint, highlighting risks related to audit trails and data governance. The document outlines a practical architecture that incorporates the Model Context Protocol (MCP) and Named Data Networking (NDN) as a data-level trust fabric. It details how these technologies can enhance security, compliance, and traceability in AI-data interactions without necessitating extensive network overhauls. The report also discusses the importance of privacy redaction and policy enforcement, emphasizing the need for enterprises to maintain control over sensitive information. Use cases in regulated industries, enterprise knowledge access, and cross-enterprise collaboration are examined, demonstrating the strategic benefits of implementing these trust layers. Overall, the report underscores the necessity of governance and auditability in AI deployments to mitigate risks and enable innovation.