Ethical guidelines for AI in medicine

The current policy briefing of the German Israeli Health Forum for Artificial Intelligence (GIHF-AI) sheds light on the use of AI in medicine with a focus on ethical principles and their influence on trustworthiness and acceptance among the population. In addition to the European Union’s ethical guidelines for trustworthy AI, Israel’s ethical framework on the use of AI in medicine will be presented as well as use cases.
Authors: Elisabeth Liebert, Lea Ledwon

The trustworthiness of an AI is essential for its acceptance by the population and its use in the European Union (EU). Therefore, any AI used in the EU should be subject to Europe-wide legal regulations such as the EU AI Act. From an ethical perspective, the reference to European fundamental values is important, as the use of AI must not harm these fundamental values. Trust is created through transparency and guidelines that both users and developers can refer to.

Both the EU guidelines and the guidelines of the German Ethics Council and ZEKO refer to similar ethical requirements. They all focus on people and their autonomy and all call for continuous quality control and the corresponding certification of the AI systems used. Another key aspect is the personal responsibility of humans: The AI systems used must not be relied on blindly. People are always ultimately responsible and should critically compare the results of the system with their knowledge. This requires the appropriate technical skills to understand how an AI works and possible sources of error within the AI. Adequate handling of the data used is also necessary, whereby the framework conditions must be enacted by the legislator.

Furthermore, what all guidelines have in common is that they are not legally binding. This raises the question of the extent to which context-specific ethical thinking can and should be translated into generally binding laws. Israel’s ethical guidelines also pave a similar path, although the issue of certification plays a subordinate role, and they tend to be more flexible and more conducive to innovation.

To balance the benefits of AI in medicine and the challenges of using the new technology, an exchange at the international level is necessary. Countries such as Israel, where the application is already more advanced, can use their experience to contribute to a constructive discourse. Regulation must identify risks and challenges, but not ignore the great benefits for medical care. Ethical use of AI also means, above all, that AI is used when it improves treatment.