Katana · AI Preferences
Katana ai preferences
Read from Katana's own page on 2026-11-09. 3 statements quoted word for word; anything the page does not say is shown as unstated, never guessed. APIs.io verifies none of it.
Train AI
stance: prohibited
Katana shall further ensure that any third-party artificial intelligence provider engaged in connection with the AI Features is contractually prohibited from using Client Data to train, fine-tune, or otherwise improve its underlying modelsRead from Katana's own page on 2026-11-09
Train generative AI
The page states nothing on this.
AI use (inference, search)
stance: allowed
In order to provide the AI Features, Katana processes prompts, uploads, and the Client Data reasonably necessary to generate a response using artificial intelligence and machine learning models through its third-party providersRead from Katana's own page on 2026-11-09
Agent onboarding
The page states nothing on this.
Agent actions
stance: prohibited
use the Katana Services or any part or element thereof by means of programs that send them automatic enquiries or requests, unless such program has been made available by Katana.Read from Katana's own page on 2026-11-09
Agent purchases
The page states nothing on this.
Published signals (probed, not read by a model)
Document
Download YAML · schema: API Commons AI Preferences 0.1.0 · transcribed by claude-haiku-5-5, checked verbatim against the fetched page