AI transparency

A fuller draft of this page, with a plain-English summary, is in our trust centre.

qypu uses AI to draft text and to generate or edit images, video and audio. This page explains how we label that content.

Human approval

Nothing is published without approval by a named person in the customer's workspace. Any change after approval requires re-approval.

Provenance (C2PA)

Media we generate carries a C2PA content-credentials manifest and IPTC 'trained algorithmic media' marking. We never strip, replace or forge provenance data, and we do not add fake camera signals.

Platform AI labels

• Instagram / Facebook: the AI-generated flag is set on AI media.
• TikTok: content is marked as AI-generated (AIGC) where it applies.
• YouTube: 'altered or synthetic content' is declared for realistic generated video.
• LinkedIn: a disclosure line is included in the post where it applies.

What we will not do

We do not offer detector evasion, fake personas, or posting to accounts customers do not control.

Models and providers

We use third-party model providers as sub-processors. [PROVIDER LIST - to publish]

Contact

Questions or reports: privacy@qypu.ai