Is the biggest problem of health-related artificial intelligence an ethical one?

Abstract

Artificial intelligence (AI) is define by MeSH (Medical Subject Headings) as “theory and
development of computer systems which perform tasks that normally require human
intelligence. Such tasks may include speech recognition, learning; visual perception;
mathematical computing; reasoning, problem solving, decision-making, and translation of
language”. The keyword has been introduced in 1986 but received lately special attention due
to the access to a considerable amount of organized/structured data. The AI technology in
health care has raw medical data as input on which apply machine learning algorithms and
provide as a specific output. The main feature of the AI is represented by the creation of its
own logic by recognizing patterns in the input data but are “black boxes” that predict well
without explaining why and are case-specific, the received goal is not self-adjusted. The healthrelated AI applications are developed to assist the diagnosis, development of the treatment
protocol, drug development, personalized medicine, and healthcare monitoring. High-Level
Expert Group on Artificial Intelligence (AI HLEG, https://ec.europa.eu/digital-singlemarket/en/high-level-expert-group-artificial-intelligence) published the Ethics Guidelines for
Trustworthy AI that listed “seven key requirements for Trustworthy AI: (1) human agency and
oversight, (2) technical robustness and safety, (3) privacy and data governance, (4) transparency,
(5) diversity, non-discrimination and fairness, (6) environmental and societal well-being and (7)
accountability.” The use of AI technology in health care and medical education arise several
ethical issues regarding patient autonomy, privacy, and confidentiality, informed consent,
discrimination, quantification of AI risks and benefits, responsibility, misuses, responsible
conduct of AI research and testing, etc. Several AI ethical issues in health-care are introduced
and discussed.