Microsoft stated that it is collaborating with over 40 organizations, including Accenture, Adobe, KPMG, MasterCard, and Volvo Cars, to promote unified transparency, security, and governance practices across different models and suppliers through a new AI transparency initiative. Microsoft will also enhance the visibility of how its products and services' AI models and agents interact with corporate data.
Redmond, Washington, USA, November 17th / PRNewswire / -- Organizations can observe a great deal of activity surrounding the AI models, but it is often difficult to reconstruct what happens between the request and the result, which makes it challenging to independently verify the behavior of these models. On Tuesday at the Microsoft Ignite conference, Satya Nadella, Microsoft's Chairman and CEO, announced the next phase of Microsoft's commitment to making AI safer: the AI Transparency Initiative. This is a collaborative effort with customers and partners aimed at establishing a common, open approach to security and governance for different models and suppliers. The initiative embodies a simple principle: the greater the AI autonomy, the higher the transparency must be.
Hayett Galott, Executive Vice President of Microsoft Security, stated: "For security controls to be truly effective, every organization needs to have greater transparency regarding the models themselves. With this effort, we will bring more transparency to all models. This is a key step in making containment measures smarter and more powerful, and it also helps us to better define responsibilities and traceability."
Enterprises are connecting the AI models to data, tools, credentials, and infrastructure, enabling agents to take action without the need for manual approval at each step. Existing control measures can observe a large amount of activity surrounding the AI models, but they cannot see what is happening inside the models, nor can they independently verify the explanations provided by the models for their own behavior. This limits an organization's ability to understand the context behind the agents' actions. Such gaps in evidence pose risks to security, compliance, auditing, and accountability. Security teams need a new trust architecture to address the growing internal risks posed by increasingly capable AI agents.
Chief Security Officer Anchar Gupta of Adobe stated: "As AI intelligents are increasingly integrated into business operations, security teams need to have a clear understanding of what these intelligents are accessing, what actions they are taking, and how those actions can be verified. Establishing a common approach to AI transparency across the industry is crucial for enhancing trust and helping organizations to innovate with confidence while maintaining security and accountability."
From Principles to Practice
The initiative is divided into three parts. First, Microsoft will expand its participation in the National Institute of Standards and Technology (NIST) in the United States and utilize its research findings to establish open, industry-wide standards for transparency, security, and governance. Second, Microsoft is collaborating with more than 40 organizations, including Accenture, Adobe, KPMG, MasterCard, and Volvo Cars, to help develop open standards that ensure that AI systems, regardless of the model or supplier used, possess characteristics of being accountable, independently verifiable, and "security by design."
Thirdly, Microsoft will incorporate this work into Microsoft AI, Microsoft Foundry, and Microsoft Copilot, enabling customers to use the resulting evidence in conjunction with their security and compliance tools, whether these tools come from Microsoft or other suppliers.
KPMG International John Israely, Chief Information Security Officer of AI, stated: "Our customers want to use AI more quickly, and they shouldn't have to make a choice between innovation and security. AI Trust requires transparency, explainability, and responsible oversight. We are delighted to collaborate with Microsoft to advance the AI transparency initiative, helping to establish open, industry-wide standards that provide organizations with the visibility and control needed for large-scale, responsible deployment of AI."
Clear accountability
Agents that have access to data, tools, and credentials have a risk profile similar to that of privileged insiders, but they operate at machine speed and scale; therefore, independent oversight is of paramount importance.
Accountability still lies with the organization itself. When major incidents occur, leaders need to be accountable to customers, employees, the board of directors, auditors, and regulatory authorities. Blaming AI is not the solution, nor is "believing in the model" a means of control.
The initiative is designed to operate across models and suppliers, including both closed models and open-weight models. As a result, development teams can replace or update the models that power their agents without compromising the ability of the IT team and security teams to govern them securely, nor will it diminish the capability of compliance teams to audit their behavior.
Harprit Sidhu, Global Head of Cybersecurity at Accenture, stated: "As organizations take the next step in their AI transformation, they need a solid foundation of trust and confidence. Microsoft's open and industry-driven approach to initiatives such as AI transparency helps us make progress in protecting AI models and implementing the controls and governance that go with them. Accenture is proud to collaborate with Microsoft and other leaders in this field."
See, verify, control
Microsoft believes that every organization should be able to see what the AI system has accessed and done, independently verify its behavior, and control the operations it is allowed to perform.
- See:Understand what tasks AI is required and authorized to perform, which models and runtimes are used, which data and tools are accessed, and what actions are taken. Continuously monitor agent activities and risk signals in accordance with policies and thresholds.
- Verification:Agents cannot be the sole source of truth for their own actions, nor can cutting-edge models be trusted to govern themselves. Evidence that covers an agent's identity, intended delegation, and actions should indicate its source (the model, the runtime surrounding it, or independent systems such as identity and strategy), in order to verify the signals reported by the model against independently observed records. Organizations should be able to use their own security and compliance systems to independently verify AI activities, rather than relying solely on the explanations provided by model providers for their behavior.
- Control:Always be aware of what AI is allowed to do. Clear evidence of the purpose, owner, data, and delegated permissions for each workflow helps an organization to implement its policies before significant actions are taken, and to reconstruct what has occurred during investigations, containment, recovery, and audits. Even as models and suppliers change, this evidence should continue to be retained by the organization. Regardless of the model or supplier, the organization should maintain control over the data that its agents can access and how they can access it.
How to Participate
Organizations interested in helping to shape this initiative or following its progress can visit aka.ms / learnmoreAItransparency to learn more and register.
Microsoft stated that by following the news on Microsoft Ignite, Microsoft and industry leaders will share insights on how organizations can confidently manage AI models and agents. For more information, please visit the Microsoft Security Blog.










