Responsible AI

Our AI principles.

The commitments that guide every Hynivo product — and that you can hold us to.

We tell you when you're dealing with AI, label AI-generated content, and explain in plain language what our systems do and their limits.

AI processing serves only what you ask for. Using your data to train models requires your separate opt-in — revocable at any time.

People stay accountable. Decisions that significantly affect individuals always have a route to human review and appeal.

We test for bias, monitor outcomes across groups and correct issues we find. We don't build social scoring or discriminatory profiling.

Layered safeguards, abuse monitoring and red-teaming help prevent misuse. We secure data with encryption and strict access controls.

We minimise the data we collect, keep it only as long as needed, never sell it and contractually bar our AI vendors from training on it.

In practice

What this means for you

Principles only matter if they change what happens. Here's how ours show up in the product and in our policies.

Clear notice before you use an AI Feature, with a link to the AI Terms
Separate switches for AI processing and AI training on our Consent page
A published list of prohibited and high-risk uses
"AI appeal" route for human review of automated outcomes
Watermarks or content credentials on generated media where supported
Regular review of our approach against laws such as the EU AI Act and UK/EU GDPR

We want to hear about harmful, inaccurate or biased output so we can fix it.