AI

Responsible AI Use

1. Our principles

2. How Solon uses AI

Solon supports tasks such as recognizing and categorizing transactions, proposing matching rules based on your transaction history, preparing reconciliation and reports, and flagging compliance deadlines. This happens largely with our own logic within your isolated environment.

3. How we use language models

For some steps, an external language model helps interpret text or suggest the best next step. In doing so:

4. Providers

We work with commercial AI providers that contractually commit not to use your requests to train their models. Where possible, we choose providers with strong data protection commitments. An overview is provided in the data processing agreement.

5. Human oversight and responsibility

AI output is a tool, not tax, legal or accounting advice. Important actions, such as entries, approvals and reports, are always submitted to a human for approval. You remain ultimately responsible for the outcomes.

6. Limitations

AI can make mistakes or be incomplete. That's why we build in checkpoints and make clear when a proposal was generated by AI, so you can review it before accepting it.

7. Oversight and regulation

We track developments in AI regulation, including the EU AI Act, and design Solon around transparency and human oversight. We update this policy whenever regulation or our use of AI changes.

8. Your control

You decide how much you let Solon do. AI-supported features are transparent and, where applicable, configurable. If you have preferences or restrictions, contact us.

9. Questions

Questions about our AI use or the safety of your data? Reach out through our contact page, read our privacy policy, or see how we secure your data.

Tutelium is offered by Tutelium, Inc. — an entity currently being formed; the registration number will follow once formation is complete.