Ethical considerations in emotion recognition technologies: a review of the literature

Katirai Amelia. Ethical considerations in emotion recognition technologies: a review of the literature. AI and Ethics 592, 167 (2023); https://doi.org/10.1007/s43681-023-00307-3.

As the global market for emotion recognition technologies (ERT)—which claim to use artificial intelligence to recognize emotions—rapidly expands, there is also increasing concern about their ethics. This paper reports on the results of a structured review of the literature on the ethics of emotion recognition technologies, to synthesize the ethical concerns expressed in the analyzed corpus of literature. This exploratory review draws on literature retrieved from the academic database Web of Science and from a hand-searching process, with a total of 43 articles included following a four-phased screening process. Three key areas of ethical concern were extracted from the literature: first, the risk of biased and unfair outcomes due to the faulty bases and problematic premises of ERT; second, the sensitivity of emotion data used by ERT; and third, the risk of harm that arises from the technologies in consequential settings, including employment, education, healthcare, and policing. This paper additionally reports on a qualitative synthesis of the guidelines for ethical use of emotion recognition technologies proposed in the literature, finding that they address the need for both ethical design and implementation, and are focused most heavily on the need for: a defined scope for the use of ERT, ethical decision-making, fairness and non-discrimination, and privacy. Ultimately, this review finds that these technologies raise significant—and potentially insurmountable—ethical issues, even as their commercial development for widespread use continues.