Trust is the cornerstone of successful AI adoption. At Microsoft, we believe that earning and maintaining user trust is what gives us permission to innovate. Our latest collaboration with Ernst & Young LLP (EY US) exemplifies this belief, highlighting our journey toward ISO/IEC 42001:2023 certification and beyond. ISO 42001 is the first certifiable and auditable standard for AI risk management.
Why Responsible AI Matters: Since its launch in November 2023, Microsoft 365 Copilot has become a widely adopted AI-powered solution, integrated into the daily workflows of millions across industries. But adoption at scale demands more than functionality—it requires trust. According to EY’s Responsible AI Pulse Survey (June 2025), while 72% of executives report integrating AI across enterprise initiatives, only one-third have implemented governance controls. This gap underscores the urgency of operationalizing responsible AI.
The EY-Microsoft Collaboration: To reinforce our commitment to responsible AI, Microsoft partnered with EY to evaluate and enhance the responsible AI practices embedded in Microsoft 365 Copilot. EY assembled a multidisciplinary team to rigorously assess our systems against ISO 42001 requirements. The result: Microsoft 365 Copilot achieved ISO 42001 certification in March 2025, becoming one of the few AI solutions globally to do so.
Key Themes from the Evaluation: EY’s evaluation surfaced five key themes that demonstrate how Microsoft operationalizes responsible AI:
- Operationalizing Policy into Practice
o Responsible AI principles are translated into actionable guidance through structured impact assessments.
o Tools like SDKs for user feedback, safety filters, and secure APIs help teams meet responsible AI requirements.
- Evaluating Harms Contextually
o Simulated harms evaluations anticipate risks like ungrounded content or jailbreak attempts.
o These evaluations are embedded into the development lifecycle to safeguard users.
- Embedding Safety Systems
o Classifiers and metaprompting shape system behavior and suppress unsafe outputs.
o EY validated the technical robustness of these layered safeguards.
- Continuous Monitoring in Production
o Metrics like uptime, accuracy, and misuse indicators feed into intelligent alerting systems.
o EY confirmed that telemetry pipelines enable rapid response to anomalies.
- Keeping Humans at the Center
o Responsible AI leads are embedded within product teams to oversee risk management.
o These champions ensure consistent governance and collaboration with Microsoft’s Office of Responsible AI.
The Trust Multiplier Effect This collaboration is more than a compliance exercise—it’s a blueprint for scalable, resilient, and adaptive responsible AI. With nearly 70% of Fortune 500 companies using Microsoft 365 Copilot, the benefits of ISO 42001 certification extend beyond Microsoft. Organizations can now accelerate their own compliance efforts by leveraging a solution that’s been rigorously tested and independently validated.
For Microsoft Responsible AI isn’t just a policy—it’s a practice. By embedding responsible AI into day-to-day workflows, continuously validating controls, and keeping humans at the center, Microsoft has built a system of accountability that evolves with technology and user expectations. This is the multiplier effect of trust: it protects, accelerates adoption, and prepares organizations to lead in an AI-driven future.
Read the full case study How Microsoft 365 Copilot Builds Customer Trust and explore how Microsoft 365 Copilot is setting the standard for responsible AI with resources below:
- Our latest Responsible AI Transparency Report that enables us to record and share our maturing practices, reflect on what we have learned, chart our goals, hold ourselves accountable, and earn the public’s trust
- Our Transparency Note for Microsoft 365 Copilot helps customers understand how our AI technology works, its capabilities and limitations, and the choices system owners can make that influence system performance and behavior
- Our Responsible AI Resources site which provides tools, practices, templates and information we believe will help many of our customers establish their responsible AI practices
Microsoft