Partager via


Code of conduct for Azure AI Speech text to speech

The following code of conduct defines the requirements that all Azure AI Speech text to speech implementations (including prebuilt neural voice, custom neural voice (pro, lite, and personal voice), prebuilt text to speech avatar, and custom text to speech avatar, video translation) must adhere to in good faith. This code of conduct is in addition to the Acceptable Use Policy in the Microsoft Product Terms.

Content requirements

  • All content released through your text to speech integration and associated metadata must be originally created by the publisher, appropriately licensed from the third-party rights holder, used as permitted by the rights holder, or used as otherwise permitted by law. It is your sole responsibility to ensure that you have appropriate rights to all content you input to the text to speech service.
  • This service must not be used to simulate the voice or image of politicians or government officials, even with their consent.

Your text to speech feature must:

  • disclose the synthetic nature of voices, images, and/or videos to users such that users are not likely to be deceived or duped—or able to prank others—into believing they are interacting with a real person; and
  • support a feedback channel that allows users of the service to report issues with the service and ensure that you address reported issues in a reasonable manner.

Your text to speech feature must not be used:

  • to deceive or intentionally misinform people;
  • for the purpose of false advertising;
  • to claim to be from any person, company, government body, or entity without explicit permission to make that representation;
  • to impersonate any person without explicit permission, including to gain information or privileges;
  • to enhance the authority or credibility of content in connection with financial, health, legal, political, or spiritual matters;
  • to create, incite, or disguise hate speech, discrimination, defamation, terrorism, or acts of violence;
  • in applications that are sexually explicit;
  • to exploit or manipulate children;
  • to make unsolicited phone calls, bulk communications, posts, or messages;
  • to disguise policy positions or political ideologies; or
  • to disseminate unattributed content or misrepresent sources.

In addition to the above, when integrating personal voice features into applications that you make available to users external to your organization, you must:

  • employ technical measures to ensure that the personal voice feature cannot be used beyond the specific use case(s) approved in advance by Microsoft via the Limited Access registration process (see Limited Access for more information);
  • employ technical measures to ensure that the feature will produce only the limited categories of output approved in advance by Microsoft via the Limited Access registration process (see Limited Access for more information);
  • allow users to opt out of using their voice and/or remove their voice from the feature at any time, through a reasonable and straightforward process;
  • implement robust security and access control measures including protecting the Azure AI Speech resource permissions, strong user authentication mechanisms, and requiring feature users to have a user account;
  • implement technical controls to ensure that users record their training data recording scripts in your application, and cannot use any pre-existing recordings to train a personal voice model;
  • collect legally sufficient consent permitting you and Microsoft to record, use, and store user’s voice data as contemplated by the terms and conditions applicable to your use of Azure AI Speech;
  • generate training data recording scripts for users to read that are dynamic and change each time a user trains their voice.
  • require each feature user to accept terms and conditions that bind them to the following before submitting any voice recordings or creating voice models:
    • each user may create personal voice models based only on their own voice;
    • users must record their training data recording scripts in your application, and may not use any pre-existing recordings to train their voice;
    • users must consent to terms satisfying all relevant legal requirements that provide adequate rights to you and Microsoft for: (a) recording and use of their voice for synthetic voice model training and use within the scope of the applicable approved use case, and (b) all associated data processing, storage, and use;
    • users must agree to give verbal consent for voice training by recording a Microsoft-provided acknowledgement statement and agreeing to terms that permit Microsoft to use such acknowledgement statement recording as a technical control to confirm that the voice in the acknowledgement statement recording matches the voice in the training data recording (the required acknowledgement statement text is available here in several languages);
    • users may use the voice model(s) they create exclusively with customer’s personal voice feature;
    • users may use the voice model(s) they create only for the specific use case(s) and categories of content approved in advance by Microsoft for the feature via customer’s Limited Access application (see Limited Access for more information);
    • users agree to using a synthetic voice based on their voice as the feature’s output voice; and
    • users agree that, among other remedies available to customer and/or Microsoft under applicable law, their access to the feature may be terminated immediately if they violate any of these terms, and the user will not be entitled to any damages or other legal relief related to termination under these circumstances.
  • inform users via clear and prominent disclosures of:
    • how the feature and personal voice work, including how you will: (a) use, process, and store recordings of their voices, (b) create voice models from their voices, and (c) generate output audio based on those voice models;
    • how they can opt out of using their voice and/or remove their voice from the feature at any time, through a reasonable and straightforward process; and
    • what users may and may not do with the feature, including a clear description of use case restrictions and limitations on output content.

Report abuse 

If you suspect that Azure AI Speech text to speech is being used in manner that is abusive or illegal, or infringes on your rights or the rights of other people, you can report it at the Report Abuse Portal.

The Microsoft Product Terms prohibit customers from using any Azure services, including text to speech, to violate the law. Customers who violate the terms of service will lose access to those services.

Changes to this code of conduct

Customers are responsible for complying with the current Code of Conduct, which may be updated from time to time. Microsoft will notify customers that have used Azure AI Speech within the previous 6 months of updates by email. If you have not used the Azure AI Speech service for more than 6 months prior to a particular update, then you will not receive a notification.