Our processing of personal data for these purposes includes both automated and manual (human) methods of processing … For example, we manually review short snippets of a small sampling of voice data we have taken steps to de-identify to improve our speech services, such as recognition and translation.
Of course, it is essential to note that without these "reviews," it would be almost impossible to refine speech recognition algorithms in order for them to understand the nuances of humans' natural speech (to understand the difference, just compare the old Apple's voice assistant vs Siri).
Microsoft assures users can control the data processed by them; however, all those promises fall a little short since they mention that such right is limited - without clarifying the scope of those limits:
In some cases, your ability to access your personal data will be limited, as required or permitted by applicable law.
Cortana, Skype, and Microsoft Support Collect, Transcribe and Share your Voice Recordings
Likewise, the guys at Microsoft updated the pages of several of their services recognizing that they collect, process and store conversations and information from their Users:When it comes to Cortana, the assistant needs to collect data for all relevant tasks:
When you use your voice to say something to Cortana or invoke skills, Microsoft uses your voice data to improve Cortana's understanding of how you speak … This may include transcription of audio recordings by Microsoft employees and vendors
Cortana collects your location periodically even if you're not interacting with the service.
Cortana also collects data about your contacts (People), such as their title, first, middle, and last names, suffix, nicknames, and company name. If you call, email, or text someone, or if they call, email, or text you, Cortana collects that person's email address or phone number.
If you want to protect yourself from Cortana, all you have left is NOT TO USE IT, and yet it will continue to work. Microsoft adds:
The Terrible Truth About Alexa
On Windows, even after you've signed out of Cortana, characters you type into the taskbar search box are automatically sent to Bing to help enable better search recommendations ... If you would prefer not to send any character data to Microsoft, you can choose not to use the search box.As of Skype, they explained that they use "transcription of audio recordings by Microsoft employees and vendors," to improve its translation features.Microsoft Support also updated its disclaimer, revealing that they collect your data and even "if you're not signed in to a voice-enabled product with your Microsoft account, your voice data might still be collected and stored," however they assure it won't be associated with your Microsoft account.