Responsible AI

Loading version…Policy Center

Our platform serves people in vulnerable circumstances. Wherever AI is used, we apply these principles, aligned with the EU AI Act and the OECD AI Principles. Each AI system below is documented in our Responsible AI assessment workbench and reviewed before launch.

Our principles

  1. No automated decisions about people. Every consequential decision is made by a trained human reviewer.
  2. Transparency. AI-generated content is labelled and you can request the original or human review.
  3. Opt-in. AI features on your content are off by default.
  4. Data minimisation. Conversation content and verification documents are not sent to third-party AI providers.
  5. Human review of moderation. AI may flag, but a human always decides.
  6. Bias monitoring. We monitor outcomes across regions, languages and communities.

Per-system assessments