Allow users to enable online speech recognition services
This `REG_DWORD` value controls, at the device level, whether users may enable Microsoft's online speech recognition services.
since 0.9.8.11
Purpose
Show registry keys (1)
HKLM\Software\Policies\Microsoft\InputPersonalization\AllowInputPersonalization
This REG_DWORD value controls, at the device level, whether users may enable Microsoft's online speech recognition services.
A value of 1 leaves the decision to each user; it does not enable online speech recognition automatically. A value of 0 disables the online services and prevents users from enabling them through Windows settings. If the value is absent, the policy is not configured and Windows also leaves the choice to the user.
When users are permitted to enable the services and choose to do so, cloud-based features such as dictation and voice input in supported applications—and Cortana on older Windows versions—can use Microsoft's online speech recognition.
Notes
Security
Online speech recognition processes voice input through Microsoft's cloud-based services. Microsoft may use voice input to improve its speech services. Setting the value to 0 prevents users from enabling those services and therefore prevents the associated submission of voice input to online speech recognition.
A value of 1 does not grant applications unrestricted microphone access. Applicable application and privacy permissions remain separate from this policy.
Possible side effects
With a value of 0, features that depend on Microsoft's cloud-based speech recognition are unavailable. These can include online dictation and voice input in dependent applications. Speech features that do not rely on this online service are not necessarily disabled.
Notes
This is a device-scoped setting mapped to Computer Configuration by Globalization.admx. Microsoft documents support beginning with Windows 10, version 1507, and later Windows versions. Starting with the Windows 10, version 1809 administrative templates, the policy uses its current online-speech-services display name; older templates called it “Allow input personalization.”
Microsoft documentation
Microsoft Learn: Privacy Policy CSP – AllowInputPersonalization
Microsoft Learn: Windows Privacy Compliance Guide