Microsoft Overhauls AI Governance in Latest Transparency Report
Microsoft has published its 2026 Responsible AI Transparency Report, which outlines the company's efforts to improve oversight and governance of artificial intelligence (AI) systems. The report marks Microsoft's third annual account of its work on AI governance and oversight.
The report highlights several key updates, including a revised version of the company's internal standards for responsible AI. These changes separate requirements for models, platform services, and applications, depending on Microsoft's role in building or deploying them.
Microsoft has also expanded its technical tools for testing and monitoring AI systems, including a new AI Red Teaming Agent and agent evaluators for measuring the quality, safety, and performance of agentic applications. The company emphasized that these changes reflect a shift towards continuous, lifecycle-based governance rather than a one-off assessment before deployment.
Another area of focus is formal certification, with Microsoft stating that it has certified against ISO 42001 across products including Microsoft 365 Copilot, Foundry, and GitHub Copilot. The company also highlighted its increased work with research bodies, standards groups, and universities to advance AI evaluation science and practice.