Ask AIRisks answers everyday questions about the risks of artificial intelligence: what they are, who should own them, what the rules say, and how to make a start. It is built for people who need to understand AI risk without becoming technical experts.

It answers only from approved, published material, such as the airisks.net reference pages and public frameworks. It doesn’t search the open internet. The full list is on the Sources page.

Who is behind it

Ask AIRisks is written and run by Human-Ai.Institute, a human-led institute. People decide what the helper may draw on; the AI only explains it.

What “sovereign” means here

The AI model, the material it reads and the questions people ask all stay on infrastructure under sovereign control. Nothing is sent to an outside AI company, and nobody’s questions are used to train someone else’s product. It runs on Boudica, a self-hosted AI platform made by OmniIndex.

Please check before you act

AI answers can be wrong or out of date. Use them to understand a subject, then check anything important with the original source or a qualified person before acting on it.

Questions, comments or a wrong answer to report? Contact us.