TUM and TUM University Hospital Secure $1.5 Million Google.org Grant to Build AIdvice Cancer Chatbot

TUM and TUM University Hospital Secure $1.5 Million Google.org Grant to Build AIdvice Cancer Chatbot

(IN BRIEF) TUM and TUM University Hospital have been awarded $1.5 million by Google.org to create AIdvice, an AI-powered chatbot that provides cancer patients and medical professionals with reliable, personalized information. Built around a bespoke guidelines database, AIdvice will use an LLM to generate responses, then break them into verifiable statements linked to original medical sources to prevent AI hallucinations. Led by Dr Jan Peeken and supported by Professor Florian Matthes, the project begins with a clinical pilot—storing data on secure hospital servers—and aims for nationwide rollout in Germany before expanding across Europe. AIdvice aspires to enhance patient understanding, streamline clinician research, and raise the overall standard of cancer information.

(PRESS RELEASE) MUNICH, 11-Jun-2025 — /EuropaWire/ — Researchers at the Technical University of Munich (TUM) and TUM University Hospital have secured a $1.5 million grant from the Google.org Accelerator: Generative AI to develop AIdvice, an AI-driven knowledge assistant tailored for cancer care. Recognizing that two-thirds of cancer patients—and even medical professionals—struggle with outdated, incomplete, or misleading online information, the AIdvice project will create a chatbot that delivers personalized, evidence-based answers in clear, accessible language.

Unlike standard large language models (LLMs) that may “hallucinate” inaccurate content, AIdvice will draw exclusively from a purpose-built knowledge database containing the latest clinical guidelines. Each response will be decomposed into individual statements, cross-verified against the database, and linked directly to original sources—such as specific sections of treatment protocols—to ensure full traceability. “Our goal isn’t merely to generate answers, but to fact-check every element against trusted guidelines,” explains Florian Matthes, Professor of Software Engineering for Business Information Systems at TUM.

Under the leadership of Dr Jan Peeken of the Clinic for Radiation Oncology and Radiotherapy, the team will build and iteratively refine the assistant over the coming years. In an initial pilot, AIdvice will be deployed in a clinical environment, with patient queries and data stored on hospital-secure servers. Once validated, the system will expand to clinics and general practices across Germany and, subsequently, other European nations—all while fully complying with healthcare data-protection regulations.

AIdvice is designed to serve both patients and their families—offering clear explanations of treatment options, side effects, and preventative measures—and to support clinicians by speeding access to up-to-date, guideline-based information. By combining advanced AI with rigorous source attribution, the project aims to combat misinformation and ultimately improve cancer care outcomes.

Further information and links

• Florian Matthes is professor for Software Engineering for Business Information Systems at the TUM School for Computation, Information and Technology.
• He is also a core member of the Munich Data Science Institute.
• Dr. Jan Peeken is a senior managing physician in the field of radiation oncology and leads a research group on the topic of artificial intelligence at the TUM University Hospital.

Media Contacts:

Corporate Communications Center
Julia Rinner
presse@tum.de