Add native voice dictation to GitHub Copilot chat

### Feature summary

Allow users to record their voice and convert speech into editable text directly within the GitHub Copilot app chat

### What problem are you trying to solve?

As a Microsoft MVP and someone who works closely with Developers, Cloud Architects, IT professionals, and technical communities, I have seen GitHub Copilot become an increasingly important part of the daily development workflow.

However, creating detailed prompts in the GitHub Copilot mobile app can still be challenging. Developers often need to provide extensive context, explain architectural decisions, describe bugs, define technical requirements, or document an idea while they are away from their computers.

Typing this level of detail on a mobile device is slow and can interrupt the natural flow of thought. In many situations, the developer already knows exactly what needs to be explained, but the friction of typing a long technical prompt discourages them from providing the full context that Copilot needs to generate a better response.

Although mobile operating systems provide general voice dictation, these solutions are not designed specifically for software development. They may struggle with technical terminology, product names, commands, file paths, programming languages, cloud services, and architecture-related vocabulary.

A native voice experience inside the GitHub Copilot app would reduce this friction and help users communicate more complete and accurate technical requirements.

Best regards.
Guilherme Maia
Microsoft Azure MVP

### Proposed solution

I would recommend adding a native microphone button directly to the GitHub Copilot chat input field.

The experience should allow users to record a voice message and convert it into editable text before submitting the prompt to Copilot.

Ideally, users should be able to:

- Start, pause, resume, cancel, and finish a voice recording.
- Convert speech into an editable transcription.
- Review and correct the text before sending it.
- Record longer technical explanations without the dictation stopping unexpectedly.
- Use automatic punctuation and paragraph organization.
- Dictate technical terms, commands, file names, product names, cloud services, and programming-related vocabulary.
- Select a language or use automatic language detection.
- Clearly understand whether transcription is processed locally or remotely.

The transcription should always be presented to the user for review before being submitted.

From a privacy and security perspective, local processing should be prioritized whenever possible. If remote processing is required, the application should clearly explain how the audio is processed, retained, and protected.

The original audio should not be permanently stored by default.

### Workflow impact

_No response_

### Installation context

_No response_

### Additional context

From my experience as a Microsoft MVP, instructor, and Technical community leader, I believe this capability could add significant value to developers who already use GitHub Copilot as part of their daily workflow.

The quality of the response generated by an AI assistant depends heavily on the quality and completeness of the context provided by the user. Voice input makes it easier to explain complex scenarios naturally, especially when describing architecture decisions, troubleshooting steps, security requirements, acceptance criteria, or implementation constraints.

This feature would be particularly useful for:

- Developers using GitHub Copilot on Mobile devices.
- Architects who need to document technical decisions quickly.
- Professionals reviewing code or repositories away from their workstations.
- Users who need to capture an idea immediately.
- People with accessibility needs or difficulty typing long prompts.
- Multilingual users who communicate more naturally through speech.

As an additional enhancement, GitHub Copilot could provide an optional “Refine voice prompt” action after transcription.

This capability could organize the spoken explanation into sections such as:

- Objective
- Current scenario
- Expected result
- Technical context
- Requirements
- Constraints
- Security considerations
- Acceptance criteria

The user should always review and approve the refined prompt before sending it.

I believe native voice dictation would make the GitHub Copilot app more productive, accessible, and useful for professionals who increasingly rely on Copilot not only for code generation, but also for planning, debugging, documentation, architecture, and technical decision-making.

No files, tests, or entry points are named in the issue. Start by scoping the supported mobile app platforms and the recording, transcription, privacy, and review requirements; done would mean an agreed implementation scope for the complete voice-to-editable-text workflow.

Rédigé par le modèle d'indexation à partir du texte de l'issue.