Microsoft and multiple companies launch the AI Transparency Initiative to enhance corporate trust in AI models and agents
PR Newswire
52m ago
Ai Focus
Microsoft announces the launch of AI Transparency Initiative, and collaborates with over 40 companies, including Accenture, Adobe, KPMG, Mastercard, and Volvo Cars, to promote common standards for transparency, security, and governance in AI. Microsoft will also integrate these capabilities into Microsoft AI, Microsoft Foundry, and Microsoft Copilot.
Helpful
No.Help

Under its new AI Transparency Initiative framework, Microsoft has collaborated with over 40 companies, including Accenture, Adobe, KPMG, Mastercard, and Volvo Cars, to promote unified transparency, security, and governance practices at the AI model and supplier levels. Microsoft also stated that it will increase visibility into how AI models and agents interact with corporate data, covering all of its products and services.

Redmond, Washington, USA, November 19, 2026 / PRNewswire / -- Enterprises can observe a great deal of activity surrounding the AI model, but it is often difficult to determine what exactly happens between a single request and its resulting outcome. This makes it challenging to independently verify the behavior of these models. On Tuesday, Satya Nadella, Microsoft's CEO, announced the next steps Microsoft is taking to make AI more secure at the Microsoft Ignite conference: AI Transparency Initiative is an effort in collaboration with customers and partners aimed at establishing a common, open approach to security and governance across models and suppliers. This initiative is based on a simple principle: the greater the autonomy, the higher the required transparency.

Microsoft's Vice President of Security, Hayete Gallot, stated: "In order 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 towards making isolation smarter and more effective, and it will also help us better define responsibilities and traceability."

Enterprises are connecting the AI models to data, tools, credentials, and infrastructure, enabling agents to act without requiring manual approval at each step. Existing control mechanisms can observe a large amount of activity around the AI models, but they are unable to see what is happening inside the models, nor can they independently verify the models' descriptions of their own behavior. This limits an organization's ability to understand the context in which agents act. This lack of evidence poses risks in terms of security, compliance, auditing, and accountability. Security teams need a new trust architecture to address the increasing internal risks posed by the increasingly powerful AI agents.

Adobe The person in charge of security, Aanchal Gupta, stated: "As AI intelligentsia become more deeply integrated into business operations, it is essential for the security team to have a clear understanding of what these intelligentsia have accessed, what actions they have taken, and how those actions can be verified. Promoting a common AI transparency approach across the industry is crucial for building trust and also helps organizations to innovate with confidence while maintaining security and accountability."

From Principles to Practice

The initiative consists of three parts. First, Microsoft will expand its participation in the efforts led by the National Institute of Standards and Technology (NIST) in the United States, and based on relevant research, develop open standards for the industry that cover transparency, security, and governance. Second, Microsoft is collaborating with more than 40 companies, including [list of companies], to promote open standards for [target systems], ensuring that these systems are accountable, independently verifiable, and secure from the outset, regardless of the model or vendor used.

Thirdly, Microsoft will integrate this work into Microsoft AI, Microsoft Foundry, and Microsoft Copilot to enable customers to use the obtained evidence in their security and compliance tools, whether these tools come from Microsoft or other suppliers.

KPMG International Global Head of Information Security John Israel stated: "Our customers wish to advance AI more quickly, but they should not have to make a choice between innovation and security. Trust in AI requires transparency, explainability, and responsible oversight. We are proud to collaborate with Microsoft to promote AI Transparency Initiative, helping to establish industry open standards that provide organizations with the visibility and control needed for responsible, large-scale deployment of AI."

Clear responsibilities

Agents that have access to data, tools, and credentials possess risk characteristics similar to those of privileged insiders, but they operate at the speed and scale of machines; therefore, independent oversight is of paramount importance.

The responsibility still lies with the enterprise itself. In the event of major incidents, management must be accountable to customers, employees, the board of directors, auditors, and regulatory authorities. Blaming AI is not a solution, nor is a "trust model" a means of control.

The initiative aims to be compatible with different models and suppliers, including both closed-weight and open-weight models. This way, development teams can replace or update the models that power their agents without compromising the ability of the IT team and security teams to manage these agents securely, nor will it weaken the compliance teams' ability to audit their behavior.

Accenture Global Cybersecurity Leader Harpreet Sidhu stated: "As organizations enter a new phase of their AI transformation, they need a solid foundation of trust. Initiatives like Microsoft's open and industry-oriented approach to AI transparency can help us protect the AI model and equip it with appropriate controls and governance. We are very 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.

  • SeeUnderstand what AI is required to do, what is allowed, which model and which runtime environment are used, which data and tools were accessed, and what actions were taken. Continuously monitor agent activities and risk signals, and conduct checks against established policies and thresholds.
  • VerificationAgents cannot be the sole source of truth for their own actions, nor can we rely on cutting-edge models to govern themselves. Evidence related to an agent's identity, intended delegation, and actions must be attributed to a specific source (the model itself, its surrounding operating environment, or independent systems such as identity management and strategy frameworks), so that the signals reported by the models can be verified against independent observations. Organizations should be capable of independently verifying AI activities using their own security and compliance systems, rather than relying solely on what model providers claim about their behavior.
  • ControlIt is essential to maintain control over what AI is allowed to do. Clear evidence regarding the objectives, owners, data, and authorization levels for each workflow helps organizations implement policies before significant actions are taken, and it also facilitates the reconstruction of events during investigation, isolation, recovery, and audit processes. Such evidence should be retained even when models and suppliers are changed. Organizations must continue to control which data their agents can access and how that access is governed, regardless of the models or suppliers being used.

How to Participate

Enterprises interested in participating in shaping this initiative or tracking its progress can visit aka.ms / learnmoreAItransparency to learn more and register.

Microsoft also stated that by following the latest developments of Microsoft Ignite, Microsoft and industry leaders will share their views on how organizations can confidently manage AI models and agents. For more information, please visit the Microsoft Security blog.

Tip
$0
Like
0
Save
0
Views 47
CoinMeta reminds readers to view blockchain rationally, stay aware of risks, and beware of virtual token issuance and speculation. All content on this site represents market information or related viewpoints only and does not constitute any form of investment advice. If you find sensitive content, please click“Report”,and we will handle it promptly。
Submit
Comment 0
Hot
Latest
No comments yet. Be the first!
Related
Ransomware hackers leaked data of over 22,000 customers from Visa, Mastercard, and JPMorgan
Ransomware gang n0n leaked customer data related to Visa, Mastercard, and JPMorgan through a third-party supplier, affecting over 22,000 people, including approximately 16,000 cardholders of JPMorgan Reserve. All three companies stated that their core systems and payment networks were not affected.
Coinpedia
·2026-11-20 04:21:43
28
Target Why did it exit NASCAR
Target withdrew from NASCAR after the end of the 2017 season, ending a 16-year sponsorship partnership with Chip Ganassi Racing. The company stated that the reason for leaving NASCAR was to shift their sports marketing budget towards football, in order to reach consumers at various levels, from professional teams to youth leagues.
Coinpedia
·2026-11-20 04:21:41
29
CoreWeave Contract Backlog Rises to $132.5 Billion, but Quarterly Interest Costs Increase to $846 Million
NVIDIA-supported AI cloud service provider CoreWeave stated that as of the end of September, its contract revenue backlog increased by 138% year-on-year to $132.5 billion, with third-quarter revenue growing by 168% to $3.66 billion. However, net losses expanded to $709 million, and quarterly net interest expenses rose to $846 million, highlighting the financing pressures brought about by its data center expansion.
Coinpaper
·2026-11-20 03:52:11
38
Anthropic Stock Price Today: Record-breaking debut on the market, once surpassing SpaceX, then fell below $100
Anthropic began trading on NASDAQ on Thursday with the code ANT, opening at $111.50, which is about 34% higher than the IPO price of $83. At the opening, the company's valuation was once around $2 trillion, briefly exceeding the record performance of its listing in June of this year SpaceX, but then the increase slowed down, and the stock price approached $99 at one point during the session.
Coinpedia
·2026-11-20 03:42:16
56
ADP Acquires ZenBusiness
ADP announces the acquisition of ZenBusiness. ADP stated that this transaction will incorporate enterprise establishment and compliance services into its product portfolio, helping entrepreneurs and small businesses manage their business, payroll, insurance, compliance, human resources, and labor management from the startup phase to expansion.
PR Newswire
·2026-11-20 03:42:14
45
View More