Skip to content

Local AI #5735

@louis-kiyus

Description

@louis-kiyus

Dupe Check

Describe the solution you'd like?

Use LLMs local to the computer to power Warp AI. None of our outputs should be sent to ANYONE, i.e. the AI provider in this case.

Is your feature request related to a problem? Please describe.

According to the privacy page, Warp only guarantees that no outputs are sent to Warp's servers, but they are still being sent to AI providers. Examining the network log shows that Warp is sending many of the output from our commands onto the Internet, and I'm really uncomfortable with that as it could very well leak sensitive informations, including private keys. Even if Warp makes an effort to redact sensitive information, it's still impossible to guarantee 100% nothing would be leaked accidentally, as sensitive data could come in many forms.

Additional context

No response

How important is this feature to you?

5 (Can't work without it!)

Warp Internal (ignore) - linear-label:39cc6478-1249-4ee7-950b-c428edfeecd1

None

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request.

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions