At Microsoft, we want to allow our customers to exploit the full potential of new technologies such as artificial intelligence, while meeting their needs and expectations of confidentiality. Today, we share the key aspects of how our approach to confidentiality protection in AI – including our concentration on security, transparency, user control and continuous compliance with protective requirements Data – are the main components of our new AI generating products like Microsoft Copilot.
We create our products with Security and confidentiality Incorporated through all the design and implementation phases. We provide transparency To allow people and organizations to understand the capacities and limits of our AI systems and the sources of information that generate the responses they receive, providing real -time information as users with our products . We provide clear tools and choices so that people can control their dataIncluding via tools to access, manage and delete personal data and store conversation history.
Our approach to privacy in AI systems is based on our longtime belief that privacy is a fundamental human right. We are committed to continuing Compliance with all applicable lawsincluding confidentiality and data protection regulations, and we support acceleration Development of appropriate railings To strengthen confidence in AI systems.
We believe that the approach we have adopted to improve privacy in our AI technology will help to clarify people on how they can control and protect their data in our new AI generator products.
Our approach
Data security is at the heart of confidentiality
Keeping secure data is an essential privacy principle in Microsoft and is essential to ensure confidence in AI systems. Microsoft implements appropriate technical and organizational measures to ensure that the data is secure and protected in our AI systems.
Microsoft has integrated Copilot into many different services, including Microsoft 365, Dynamics 365, Viva Sales and Power Platform: Each product is created and deployed with critical security, conformity and confidentiality policies and processes. Our security and confidentiality teams use both confidentiality and security by design throughout the development and deployment of all our products. We use several layers of protective measures to keep data secure in our AI products such as Microsoft Copilot, including technical controls such as encryption, which all play a crucial role in data security in our AI systems. Keep protected and secure data in AI systems – and guarantee that systems are architects to respect data access and processing policies – are at the heart of our approach. Security and confidentiality are principles integrated into our internal standard responsible for AI and we are committed to continuing to focus on confidentiality and security to keep our AI products in safety and trustworthy.
Transparency
Transparency is another key principle for the integration of AI into Microsoft products and services in a way that promotes user control and confidentiality, and strengthens trust. This is why we are committed to building transparency in people’s interactions with our AI systems. This approach to transparency begins with user clarity when they interact with an AI system if there is a risk that they are confused. And we provide real -time information to help people better understand the functioning of AI’s features.
Microsoft Copilot uses a variety of transparency approaches that meet the users where they are. Copilot provides clear information on how it collects and uses data, as well as its capacities and limits. Our approach to transparency also helps people understand how they can better take advantage of Copilot’s abilities as a daily AI tool and offer opportunities to find out more and provide comments.
Transparent choice and disclosure while users engage with Microsoft Copilot
To help people understand the capacities of these new AI tools, Copilot provides information in products that clearly allow users to know that they interact with AI and provide easy -to -understand choices in a conversational style. As people interact, these disclosure and choices help better understand how to exploit the advantages of AI and limit potential risks.
Earthly evidence and sources
Copilot also provides information on how his answers are centered or “to the earth”, on the relevant content. In our AI offers in Bing, Copilot.microsoft.com, Microsoft Edge and Windows, our Copilot answers include information on the content of the web that helped generate the answer. In Copilot for Microsoft 365, responses may also include information on user’s commercial data included in a generated response, such as emails or documents to which you already have access to access. By sharing links to entry sources and source materials, people have greater control of their IA experience and can better assess the credibility and relevance of the outputs of Microsoft Copilot and access more information to need.
Data protection user checks
Microsoft provides tools that put people to control their data. We believe that all organizations offering AI technology should guarantee that consumers can significantly exercise their subject rights.
Microsoft offers the possibility of controlling your interactions with Microsoft products and services and honors your privacy choices. Thanks to the Microsoft Privacy dashboard, our account holders can access, manage and delete their personal data and store conversation history. In Microsoft Copilot, we honor additional confidentiality choices that our users have made in our cookie banners and other controls, including choices concerning the collection and use of data.
The Microsoft Privacy dashboard allows users to access, manage and delete their data when connected to their Microsoft account
Additional transparency On our privacy practices
Microsoft provides more in -depth information on how we protect the confidentiality of individuals in Microsoft Copilot and our other AI products in our transparency materials such as M365 COPILOT FAQS And The New Bing: Our AIA AIR AIAwhich are accessible to the public online. These transparency materials describe in more detail how our AI products are designed, tested and deployed – and how our AI products approach ethical and social problems, such as equity, confidentiality, security and responsibility. Our users and the public can also examine the Microsoft confidentiality declaration which provides information on our privacy practices and controls for all Microsoft consumer products.
AI systems are new and complex, and we always learn how we can better inform our users of our new AI revolutionary tools significantly. We continue to listen to and incorporate comments to make sure that we provide clear information on the operation of Microsoft Copilot.
Comply with current laws and the support of progress in the global data protection regulations
Microsoft is in accordance with data protection laws today in all the courts where we operate. We will continue to work in close collaboration with governments around the world to make sure to remain in conformity, even if the legal requirements develop and change.
Companies that develop AI systems have an important role to play in work with regulators of confidentiality and data protection from around the world to help them understand how AI technology is evolving. We are committed to regulators to share information on the operation of our AI systems, how they protect personal data, the lessons we have learned because we have developed the confidentiality, security and governance systems of AA Responsible, and our ideas on how to solve the unique problems concerning AI and confidentiality.
AI regulatory approaches are advancing in the European Union through its Ai acand in the United States through the president Executive decree. We expect additional regulators from around the world will seek to meet the opportunities and challenges that new IA technologies will bring to privacy and other fundamental rights. Microsoft’s contribution to this global regulatory discussion includes our plan to govern AI, where we make suggestions on the variety of approaches and controls that governments may want to consider to protect privacy, advance fundamental rights and guarantee that AI systems are safe. We will continue to work closely with data protection authorities and confidentiality regulators around the world as they develop their approaches.
As the company progresses in this AI era, we will need the leaders of privacy within the government, organizations, civil society and the academic world to work together to advance harmonized regulations that guarantee to AI innovations to all and focus on the protection of privacy and other fundamental human rights.
At Microsoft, we are committed to doing our part.