When algorithms are involved in diagnosing diseases and making recommendations for treatments, ethics are required. A year ago, the German Ethics Council, led by Munich medical ethicist Alena Buyx, developed a detailed statement on the topic of “Humans and Machines – Challenges posed by Artificial Intelligence”. During the health forum organized by the Süddeutsche Zeitung and the Bavarian Academy of Sciences on the topic of AI in medicine, Buyx reported on the main conclusions. According to the Chairwoman of the Ethics Council, AI is without question a dual-use technology, a mixture of “horror and savior”. Both aspects are often exaggerated in the public debate, says Buyx: “The truth lies in the middle.”
AI offers “fantastic possibilities, especially in medicine,” said the ethicist. It could revolutionize the diagnosis of skin cancer, improve mammography and X-ray analyzes and also help in the targeted development of medications. “Now it’s about using this technology responsibly with the greatest benefit.”
Digital legislation in the healthcare system is overdue
As intelligent as technology is, AI cannot match human intelligence in all its dimensions, said Buyx. “We do not deny that the cognitive and computational areas of intelligence can soon be taken over by these machines. At the same time, we clearly emphasize: the human intelligence that controls our planned, responsible actions is more than just cognitive and computational, it is also emotional, emotive and social. This is more than just data collection and wiring.” For this reason, AI applications cannot assume responsibility in a morally relevant sense. “The use of AI must expand human development, authorship and possibilities for action and must not diminish them,” said Buyx. “AI must not replace humans.” Under no circumstances should doctors be obliged to use the results of AI. But the next generation of doctors must be well trained to use the technology for the benefit of patients – and they must be informed about its use.
As an ethicist, Buyx emphasized that she feels committed to data protection. But it is also clear: “Good AI only works with good data. If we want the healthcare system to stay good and get better, then we should all work together to promote the common good-oriented use of medical data, responsibly and securely.” She is therefore “despite all the debate and all the criticism, satisfied that some digital legislation is now being passed in the healthcare system”.