In a recent ruling from 12 May 2026, the German company Aesthetify GmbH was held liable for the hallucinations of their company chatbot. The Hamm Higher Regional Court ruled that a company can be held responsible for false information generated by its AI chatbot even when the chatbot was trained using accurate data.
Patients would interact with the chatbot to ask questions and book appointments online, relying on the accuracy of the information presented to make an informed medical choice. However, the chatbot incorrectly described two doctors as holding specialist medical titles that did not exist.
Although Aesthetify argued that it should not be responsible for inaccurate answers produced autonomously by the chatbot, the court disagreed and confirmed that the company was liable for the output of its own chatbot - the chatbot could not be seen as a third party in this case.
This legal case highlights the growing complexity of governance in the AI field. It is becoming more clear with recent rulings that deploying AI is not a path for reducing organisational responsibility - if anything, the addition of AI services increases the need for human oversight as well as testing and accountability frameworks.
💭 If an AI chatbot gives false or harmful information, should responsibility always remain with the company deploying it?
