Good Healthcare AI Knows When Not to Answer

Hungarian healthcare publication Weborvos interviewed MedHubAI founders Bence Csernák and Tamás Csernák about Anna AI, a digital support tool developed for people living with cancer. The conversation focuses on a question every healthcare AI team must answer: where should the system stop, and when should a human professional take over?

What the interview covers

Anna AI grew from the oncopsychology knowledge and clinical experience of Dr Ágnes Riskó. The team describes a development process combining medically reviewed material, defined limits and repeated testing against real questions and feedback. The goal is focused support within a clearly stated role, rather than a general chatbot making claims across healthcare.

• Defined limits: Anna AI provides information, education and emotional support, but it does not provide psychotherapy or medical advice.
• Human handover: when a question falls outside its role or suggests that specialist support is needed, the system directs the person to an appropriate professional or support service.
• Reviewed knowledge: the system is built around focused, professionally reviewed material instead of relying on a general-purpose model alone.
• Hybrid care: the team expects AI to improve access to basic support and guidance while personal therapeutic relationships remain essential.

Read the full interview in Hungarian on Weborvos →

This page is a short English summary. The complete interview and its full context are available from the original publisher.