Description: A new and dystopian study published by researchers at University of Kentucky and Indiana University touts the benefits of an AI algorithm that can purportedly diagnose depression in CEOs by analyzing their vocal features captured during earnings call recordings. The authors of the study note, “Despite its high prevalence, CEO depression is often concealed due to stigma and concerns about professional reputation, making it underrecognized and undertreated.” And yet, despite acknowledging the stigma and risks of disclosure, the researchers are touting an intrusive and dangerous surveillance tool that can potentially be used to “out” CEOs who are experiencing mental illness.

The author of the Fortune article, Sasha Rogelberg, who is merely reporting on the study, implies that the incentives to use such tools are not based primarily on concerns about executives’ well-being, but rather the potential to read into the CEO’s psychology (think new form of insider trading) or even to justify removal of CEOs whose companies are underperforming. The authors of the study even note that “Of course, it is important for the personal health of these executives, but also has far-reaching implications for the organization, the employees, the investors, and the broader economy.” Importantly, Rogelberg adds that “Eight in 10 CEOs and 67% of employees believe someone with a mental illness is weak or burdensome.”

One critical problem with the study’s results is that we would expect some correlation between a firm experiencing negative performance, including “litigation or volatile stock returns,” and the CEO betraying a certain amount of negative affect in their voice. Although there is no reason to doubt a high prevalence of depression among corporate executives especially when their firms are going through difficult times, it’s ethically dubious to suggest that it might benefit employees and executives to start diagnosing them without their knowledge or consent or to suggest that such diagnostic tools are even valid. I think it would have been reasonable for Rogelberg to do additional reporting for this article by seeking commentary from bioethics scholars, algorithmic researchers and activists, and mental health advocates.

Source: Fortune