1. When you are interacting with AI
When you chat with Aria, or use AI-assisted classification or document drafting, you are interacting with an AI system — not a human and not a lawyer. The product labels these features as AI-assisted so you always know.
2. What the AI does
Our AI features help you work faster: Aria answers questions about the EU AI Act and your systems in context, suggests draft text for documents, and helps surface relevant obligations. Everything it produces is a starting point for your review.
3. The model behind Aria
Aria is powered by a third-party large language model — currently Mistral, hosted in the EU. We may change model providers to maintain or improve the service and keep our sub-processor list and Data Processing Agreement current. A fuller technical overview — the model, sources and testing — is published at How Aria is built and tested.
4. What the AI does not do
AI responses are informational only. They may contain errors, omissions, hallucinations, or outdated information. They are not legal advice, nota regulatory determination of compliance, and do not create any safe harbour with any authority. Always verify a factual or legal claim before acting on it.
5. Human oversight
The service is designed around human-in-the-loop use: a competent person reviews, edits, and approves AI output before it is relied upon. There is no automated decision-making that produces legal effects without your review. The product surfaces, and records, point-of-use acknowledgements when you first use Aria, export a document, or mark an obligation complete.
6. Your data and AI
We do not use your content to train foundation models. Your inputs to Aria are processed to generate a response and handled under our Privacy Policy and Data Processing Agreement. Avoid entering special-category personal data — the service is not designed to hold it.
7. Contact
Veritome by Relay Labs Ltd.
Email: info@veritome.eu
Dublin, Ireland