Understanding AI limitations
Hallucination risk, verification, and prototype features.
AI agents reason over models and retrieved context. They can be wrong — especially when knowledge is missing, integrations are not connected, or the scenario is illustrative rather than live data.
Verify before you act
- Treat proposed fixes, rollbacks, and customer quotes as drafts until a human validates them.
- For prices, availability, and bookings — always confirm against connected sources.
- The homepage Dave Doctor demo is simulated — not evidence of live system access.
Next steps
Read FAQ and Business knowledge and FAQs.