Automating medicine the ethical way

Abstract

Greatly increased automation in medicine is probably inevitable. Artificial Intelligence (AI) is being introduced into medical practice is numerous ways from machine diagnosis, through robot surgery to psychotherapy that is delivered entirely by computer. However, there are many ethical challenges involved. The emerging field of machine medical ethics currently lags behind medical practice and contains many unresolved debates. In aviation a very high level of automation has brought increased safety and efficiency. However, this was not achieved by simply building the technology and requiring professionals to use it. Lessons from the aviation industry suggest that issues of acceptance and resistance by professionals can be successfully handled if they are fully engaged in the operational and procedural changes at all stages. Negotiation over procedures and responsibility for errors in aviation is complex and informative for other fields. If the aviation industry exemplifies the potential benefits, then the Information Technology (IT) industry, particularly AI, highlights the dangers. Since this industry produces the ‘machine’ part of machine medical ethics, it is unfortunately necessary to observe that, historically the IT industry has demonstrated rather low standards of ethics and social responsibility. Improved ethical awareness and professionalism will be needed for IT professionals to achieve ethically acceptable technology in medicine. Long-standing unresolved debates in IT ethics now need some conclusion. These include the ethical responsibility of IT professionals for unreliable technology or for human errors resulting from poor user interfaces. Given the pace of progress in both technology and clinical practice, there is an urgent need for progress in machine medical ethics. A brief set of initial recommendations is provided.