Incident file · Security
Microsoft 365 Copilot "oversharing" (2024, ongoing)
latent permission debt made discoverable at scale, delayed 40% of surveyed rollouts 3+ months
The system
Microsoft 365 Copilot is an enterprise AI assistant grounded over an organization's own SharePoint and OneDrive content through what Microsoft calls its "Semantic Index". Microsoft's documentation states Copilot "presents only data that each individual can access" using the organization's existing permission model: Copilot is designed to respect the identity-based access boundary, not to bypass it.
What was observed
As organizations rolled out Copilot, many discovered it could surface payroll data, confidential legal files, and strategy documents to end users who technically had "view" access but, in practice, had never found or seen those files before Copilot existed.
Stop and think
- If a user technically had "view" access to a file before Copilot, and Copilot only shows files a user has access to, who or what is actually responsible for the exposure?
- What is the difference between "has access to a file" and "can easily discover a file by typing a natural-language question"? Are those the same risk?
- If permission hygiene in your document store has been accumulating for years, what would it take to find out how bad it already is, before you turn on an AI search layer?
Answer the questions above before you read on: what failed, why, and what would have stopped it.