Organizations throughout industries are leveraging Microsoft Azure OpenAI Service and Copilot companies and capabilities to drive progress, enhance productiveness, and create value-added experiences. From advancing medical breakthroughs to streamlining manufacturing operations, our prospects belief that their knowledge is protected by sturdy privateness protections and knowledge governance practices. As our prospects proceed to broaden their use of our AI options, they are often assured that their useful knowledge is safeguarded by industry-leading knowledge governance and privateness practices in probably the most trusted cloud in the marketplace right this moment.
At Microsoft, we have now a long-standing observe of defending our prospects’ data. Our method to Accountable AI is constructed on a basis of privateness, and we stay devoted to upholding core values of privateness, safety, and security in all our generative AI merchandise and options.
Microsoft’s current privateness commitments lengthen to our AI industrial merchandise
Business and public sector prospects can relaxation assured that the privateness commitments they’ve lengthy relied on for our enterprise cloud merchandise additionally apply to our enterprise generative AI options, together with Azure OpenAI Service and our Copilots.
- You might be answerable for your group’s knowledge. Your knowledge is just not utilized in undisclosed methods or with out your permission. Chances are you’ll select to customise your use of Azure OpenAI Service by opting to make use of your knowledge to wonderful tune fashions on your group’s personal use. Should you do use your group’s knowledge to wonderful tune, any fine-tuned AI options created together with your knowledge will probably be obtainable solely to you.
- Your entry management and enterprise insurance policies are maintained. To guard privateness inside your group when utilizing enterprise merchandise with generative AI capabilities, your current permissions and entry controls will proceed to use to make sure that your group’s knowledge is displayed solely to these customers to whom you have got given acceptable permissions.
- Your group’s knowledge is just not shared. Microsoft doesn’t share your knowledge with third events with out your permission. Your knowledge, together with the information generated by means of your group’s use of Azure OpenAI Service or Copilots – resembling prompts and responses – are stored non-public and will not be disclosed to 3rd events.
- Your group’s knowledge privateness and safety are protected by design. Safety and privateness are integrated by means of all phases of design and implementation of Azure OpenAI Service and Copilots. As with all our merchandise, we offer a powerful privateness and safety baseline and make obtainable further protections that you could select to allow. As exterior threats evolve, we are going to proceed to advance our options and choices to make sure world-class privateness and safety in Azure OpenAI Service and Copilots, and we are going to proceed to be clear about our method.
- Your group’s knowledge is just not used to coach basis fashions. Microsoft’s generative AI options, together with Azure OpenAI Service and Copilot companies and capabilities, don’t use your group’s knowledge to coach basis fashions with out your permission. Your knowledge is just not obtainable to OpenAI or used to coach OpenAI fashions.
- Our merchandise and options proceed to adjust to world knowledge safety laws. The Microsoft AI merchandise and options you deploy proceed to be compliant with right this moment’s world knowledge safety and privateness laws. As we proceed to navigate the way forward for AI collectively, together with the implementation of the EU AI Act and different legal guidelines globally, organizations could be sure that Microsoft will probably be clear about our privateness, security, and safety practices. We’ll adjust to legal guidelines globally that govern AI, and again up our guarantees with clear contractual commitments.
You will discover further particulars about how Microsoft’s privateness commitments apply to Azure OpenAI and Copilots right here.
We offer applications, transparency documentation, and instruments to help your AI deployment
To help our prospects and empower their use of AI, Microsoft gives a spread of options, tooling, and assets to help of their AI deployment, from complete transparency documentation to a set of instruments for knowledge governance, threat, and compliance. Devoted applications resembling our industry-leading AI Assurance program and Buyer Copyright Dedication additional broaden the help we provide industrial prospects in addressing their wants.
Microsoft’s AI Assurance Program helps prospects be sure that the AI functions they deploy on our platforms meet the authorized and regulatory necessities for accountable AI. This system contains help for regulatory engagement and advocacy, threat framework implementation and the creation of a buyer council.
For many years we’ve defended our prospects in opposition to mental property claims referring to our merchandise. Constructing on our earlier AI buyer commitments, Microsoft introduced our Buyer Copyright Dedication, which extends our mental property indemnity help to each our industrial Copilot companies and our Azure OpenAI Service. Now, if a 3rd occasion sues a industrial buyer for copyright infringement for utilizing Microsoft’s Copilots or Azure OpenAI Service, or for the output they generate, we are going to defend the shopper and pay the quantity of any adversarial judgments or settlements that outcome from the lawsuit, so long as the shopper has used the guardrails and content material filters we have now constructed into our merchandise.
Our complete transparency documentation about Azure OpenAI Service and Copilot and the shopper instruments we offer assist organizations perceive how our AI merchandise work and present decisions our prospects can use to affect system efficiency and habits.
Azure’s enterprise-grade protections present a powerful basis upon which prospects can construct their knowledge privateness, safety, and compliance programs to confidently scale AI whereas managing threat and guaranteeing compliance. With a spread of options within the Microsoft Purview household of merchandise, organizations can additional uncover, defend, and govern their knowledge when utilizing Copilot for Microsoft 365 inside their organizations.
With Microsoft Purview, prospects can uncover dangers related to knowledge and customers, resembling which prompts embrace delicate knowledge. They will defend that delicate knowledge with sensitivity labels and classifications, which suggests Copilot will solely summarize content material for customers once they have the correct permissions to the content material. And when delicate knowledge is included in a Copilot immediate, the Copilot generated output mechanically inherits the label from the reference file. Equally, if a person asks Copilot to create new content material primarily based on a labeled doc, the Copilot generated output mechanically inherits the sensitivity label together with all its safety, like knowledge loss prevention insurance policies.
Copilot dialog inherits sensitivity label
Lastly, our prospects can govern their Copilot utilization to adjust to regulatory and code of conduct insurance policies by means of audit logging, eDiscovery, knowledge lifecycle administration, and machine-learning primarily based detection of coverage violations.
As we proceed to innovate and supply new sorts of AI options, Microsoft will proceed to supply industry-leading instruments, transparency assets, and help for our prospects of their AI journey, and stay steadfast in defending our prospects’ knowledge.
