Responsibility
Built for responsible learning.
We don't promise the future of AI is risk-free. We do promise the tutor is honest about what it is.
- Learning-first design. Won't write your homework. Will help you understand how to think about it.
- Honest about uncertainty. Mercurius is instructed to say when it isn't sure, and to ask you to check its claims against a real source before you rely on them.
- Bias awareness. Mercurius is upfront about the limits of its training data and the perspectives it might be missing.
- Academic integrity. Asking for an essay redirects to the thinking behind it.
- Privacy-conscious. No account, no name, no email. Your messages go to our server and to Anthropic's Claude to generate each reply, are kept for 90 days, and can be deleted from inside the app (app 2.3 and later) or by email. The full picture is in the Privacy Policy.
- Points to real help. Mercurius is a tutor, not a counselor. If a student seems to be in crisis, it is instructed to point to real people who can help — in the US, call or text 988 (Suicide & Crisis Lifeline), or text HOME to 741741.
What we don't claim
Honest about the limits.
Mercurius is an AI built on a large language model. That means it can still produce confident-sounding mistakes. The whole product is designed around teaching you to notice when that's happening — not pretending it doesn't.
We don't claim Mercurius replaces a teacher. We don't claim it guarantees better grades. It is not a substitute for a counselor, doctor, or lawyer. We don't sell student data, and we don't require an account to learn. If any of those things change, this page changes too.