About cloud-based AI voice features

Last updated on Sep 24, 2026

Learn how cloud-based AI voice features analyze and improve dialogue in Adobe Premiere, and what you agree to when you use them.

Editing dialogue-heavy footage like interviews, panels, and voiceovers usually means hours of manual audio cleanup. Premiere's cloud-based AI voice features use Adobe's AI models to speed up the process by analyzing and improving spoken audio right in your project.

What are cloud-based AI voice features?

Cloud-based AI voice features analyze voices in the files you work with to provide audio-editing capabilities. When you use a cloud-based AI voice feature, Premiere uploads the relevant audio to Adobe for processing with cloud-based AI models and returns the result to your project.

You must be signed in and connected to the internet to use these features. Each cloud-based AI voice feature may process voices differently. 

Separate Crosstalk

Separate Crosstalk separates overlapping dialogue from multiple speakers into individual audio tracks. It can be useful when editing interviews, panel discussions, documentary footage, or other recordings where people speak at the same time. After the voices are separated, you can edit, mix, and process each audio track independently.

What does Separate Crosstalk process, and what do I consent to?

When you use Separate Crosstalk, Premiere sends the selected audio to Adobe’s cloud-based AI model. The model analyzes characteristics of the voices in the audio to separate overlapping speakers into individual tracks. This may involve creating and processing biometric information, including numerical representations of voice characteristics called voice embeddings.

Adobe uses this information only to provide Separate Crosstalk. It isn't stored beyond providing the feature, and Adobe does not use your content to train AI models.

Before using Separate Crosstalk for the first time, you'll be asked to authorize Adobe to process the information in your uploaded files for these purposes. You must also confirm that you have all necessary rights and permissions from the people featured in your files. After processing is complete, the separated audio is returned to your Premiere project. 

Note

Separate Crosstalk isn't available on files uploaded from Illinois or Texas due to legal restrictions.

What happens if I do not provide consent?

You cannot use Separate Crosstalk if you don't provide consent. You can stop using Separate Crosstalk at any time. Learn about the Adobe Privacy Policy for details on privacy rights and requests concerning personal information associated with your account.

Frequently asked questions

No, these features use cloud-based AI models. You must be signed in to Premiere and connected to the internet to use them.

Separate Crosstalk is currently the only cloud-based AI voice feature available in Premiere. Additional features within scope will be added over time.