How to Roll Out Microsoft Copilot Securely and Ensure Data Security

• By Farrah Gamboa
• Mar 21, 2024

Microsoft Copilot is a robust AI productivity tool integrated into Microsoft 365 applications. Copilot can significantly improve your daily workflows by assisting you with drafting documents and presentations, capturing action items in Teams meetings, analyzing data in Excel, and other tasks.

However, adopting Copilot can also introduce unexpected data security risks because it is built on native access controls within Microsoft 365, meaning it can access all data a user has access to, including documents, emails, and notes. Therefore, it is crucial to prepare your organization for a secure Copilot rollout and ensure that your sensitive data remains under control in the future.

Understanding the Risks Involved in a Copilot Rollout

The key factors that make Copilot adoption risky include:

• Improper permissions — Copilot relies on the permissions assigned in Microsoft 365. If users or groups have been granted inappropriate access to content, then sensitive information can quickly spiral out of control.
• Inaccurate data classification — Copilot is also governed by the sensitivity labels assigned to protect data. Therefore, if those labels are not accurate, data is at risk. Unfortunately, data classification is often inconsistent and incomplete. For example, manual labeling is highly prone to human errors and not scalable to the massive volume of content that organizations typically have, and Microsoft labeling technology is limited to specific types of files.
• Copilot-generated content — New documents generated by Copilot do not inherit any sensitivity labels from the source documents. As a result, new documents containing sensitive data could be shared with unauthorized users. But ensuring these documents are appropriately classified is a huge challenge due to the sheer volume of content that Copilot can produce.

Mitigating the Risks of Breaches when Using Copilot

If these risks are not addressed, adopting Copilot exposes organizations to data breaches and compliance fines. According to the latest Ponemon Institute's Cost of Insider Risks report, the average insider threat cost rose to $16.2m per organization in 2023, up from $15.4m in 2022. Therefore, it is crucial to enhance your data security posture before implementing Copilot and to establish a robust data access governance program to ensure continued security post-implementation.

The following capabilities enable organizations to address all three risk factors associated with the adoption and ongoing use of Copilot:

• Implementing and maintaining a strict least-privilege model helps ensure that users have just enough permissions to perform their job duties, which reduces the risk of data breaches. The process must include easy entitlement reviews by data owners, access request and approval workflows, and visibility into adequate access.
• Automated data discovery and classification can help ensure consistent and accurate labeling of both existing data stores and new content generated by Copilot. As a result, implementing appropriate security controls around all content becomes much less challenging. Furthermore, accurate data labeling is crucial in establishing an effective data loss prevention (DLP) strategy.
• Automated risk remediation detects conditions that put valuable data at risk and automatically remediates them. For example, by proactively revoking excessive permissions and disabling user accounts that are behaving suspiciously, organizations prevent sensitive information from being compromised.
• Alerts on threats, such as attempts to change permissions or read a sensitive document, empower the security team to investigate promptly and respond effectively to avoid or at least minimize damage.

By implementing these capabilities, you can confidently roll out Microsoft Copilot and reduce the risk of data breaches after adoption.