Responsible AI

This document sets out Microsoft's rules and values for how it builds and deploys AI responsibly, including commitments to fairness, privacy, and safety. When you use Copilot at work, it follows your organization's existing security rules so only people already authorized can see what it produces. You can manage your own privacy preferences at any time through your Copilot Privacy settings.

This document establishes Microsoft's responsible AI framework, grounding the development and deployment of AI systems in six enumerated core values: Fairness, Reliability and Safety, Privacy and Security, Transparency, Accountability, and Inclusiveness. It mandates the use of AI governance tooling—specifically the Responsible AI Dashboard—to monitor and manage AI systems, and assigns institutional oversight to a dedicated Office of Responsible AI. The framework places legal and regulatory compliance as a named obligation within responsible AI use, requires responsible validation of AI models for fairness and real-world alignment, and establishes that Copilot operates within and inherits the user's existing security and compliance permissions rather than creating a new security perimeter. Users retain persistent control over privacy preferences through Copilot Privacy settings.

For individual users, this document means that Microsoft's AI systems are developed under a framework designed to prevent discrimination and promote fairness, and that Copilot at work does not bypass your organization's existing access controls. Your privacy preferences are not fixed—you can manage them at any time through Copilot Privacy settings, giving you a persistent, user-controlled mechanism to adjust how your data is handled.