Artificial Intelligence
Written by: Tathagata Sen
Updated 11:38 PM EST, November 20, 2026

Microsoft has launched an initiative with more than 40 organizations to develop open standards for AI transparency, security and governance across models and providers, according to its November 17 announcement at Microsoft Ignite in Redmond, Washington.
The effort aims to help enterprises independently verify AI agents’ access and actions.
Participants include Accenture, Adobe, KPMG, Mastercard and Volvo Cars. Microsoft will also expand its participation in the National Institute of Standards and Technology’s (NIST) super intelligence (SI) Agent Standards Initiative and plans to incorporate the work into Microsoft AI, Foundry and Copilot.
Microsoft’s approach centers on three capabilities: seeing, verifying and controlling AI activity.
This involves recording an agent’s instructions, permissions, model, data access and actions; independently verifying that evidence; and retaining control over its permissions and delegated authority.
Microsoft also wants these records to remain usable when organizations switch AI models or providers.
The initiative has implications for chief data officers (CDOs), who must ensure AI agents access enterprise data appropriately and remain accountable for how they use it.
This requires clear records of which data agents can access, who authorized that access and what actions they perform. CDOs will need to coordinate data governance with AI oversight, identity management and cybersecurity, while establishing controls that work across AI platforms.
These priorities also reinforce the need for robust AI governance programs.
The initiative remains at an early stage. Microsoft has not yet announced a technical specification, common audit-log format or implementation timeline.