Ethical Concerns of AI Implementation in Computer Based Mental Healthcare

Sammanfattning

As the issues surrounding what has come to be referred to as the “global mental health pandemic” only seem to increase in severity for every passing year, medical researchers and practitioners are constantly looking to alleviate the situation by levying modern advancements in technology and medical knowledge. The promise of increased efficiency and enhanced predictive and diagnostic capabilities of medical systems powered by artificial intelligence is understandably seen as incredibly promising, enabling streamlined treatments and more accurate diagnoses for complex cases. However, the many positives of these systems bring with them a suite of very real and present negatives, not least of all concerns regarding the ethics of relinquishing one’s mental health to the mercy of a machine. This thesis is a systematic qualitative literature review, and utilizes the research question “What are the ethical considerations and potential risks associated with using artificial intelligence and machine learning algorithms to assess and treat mental health conditions in modern mental healthcare?” as an analytical foci aimed at facilitating a closing of the current knowledge gap related to how these ethical considerations influence the usage and implementation of AI-powered medical systems. Through purposive sampling through analysis of abstracts and pertinent chapters in contemporary research literature in the medical field, 19 articles and publications were studied through thematic analysis to yield the primary study results. The nascence of most derivative AI technologies currently employed in the field of mental healthcare led to the most notable ethical considerations lacking significant distinction from the ethical considerations of AI in general terms, such as inherent bias and privacy concerns, although some medically specific particularities were discovered as well.