Responsible AI

This document sets out Microsoft's rules and principles for how it builds and deploys AI responsibly, covering fairness, privacy, safety, and accountability. When you use Copilot at work, it operates within your organization's existing security permissions, so only people already authorized can access what it generates. You can manage your privacy preferences at any time through your Copilot Privacy settings.

This document establishes Microsoft's responsible AI governance framework, organized around six core values: Fairness, Reliability and Safety, Privacy and Security, Transparency, Accountability, and Inclusiveness. It sets out substantive obligations for AI development and deployment, including non-discrimination requirements, bias mitigation, mandatory implementation of AI governance tooling such as the Responsible AI Dashboard, and the inheritance of existing security and compliance frameworks when Copilot is used in enterprise contexts. The document assigns institutional responsibility for ethics and governance oversight to a dedicated Office of Responsible AI, and identifies legal and regulatory compliance as a named obligation within the responsible AI framework. Users retain persistent access to privacy preference controls through Copilot Privacy settings.

This document means that Microsoft's AI systems are governed by stated principles requiring equal treatment and non-discrimination, and that the ethical framework applied to generative AI tools explicitly includes bias mitigation, privacy, security, and reliability. When using Copilot at work, your existing organizational security permissions control who can see the content it produces — Copilot does not create a separate or broader access layer. As a concrete action available to you: you can manage your privacy preferences at any time by visiting your Copilot Privacy settings.