With artificial intelligence (AI) becoming omnipresent in our lives today, it’s no surprise that chatbots similarly are becoming a major presence across the health care system. With their emergence, many health equity advocates are digging deeper in analyzing whether chatbots foster bias.
In a resent study conducted by University of Colorado School of Medicine, the key focus was on the experience of patients with AI and, specifically, whether chatbots enhance health outcomes for patients. Given that chatbots often are the first interaction that many patients have with the health care system, the researchers determined that it is essential to best understand the potential effects of chatbots on patients’ trust and feelings of security.
The researchers focused on how individuals observe and identify the race or ethnicity of a chatbot and, in turn, how that perception might impact their experience. For example, might patients share more with chatbots they perceive to be of the same ethnicity, race, gender, or other characteristics?
The study found that chatbots can, in fact, lead to bias and greater health inequities if the nature and characteristics of a chatbox perpetuate a level of lack of trust and security that is similar to the patient’s human experience. By contrast, intentional design can also potentially facilitate enhanced trust in health care workers, medical products and overall health systems, if done effectively.
As chatbots, and AI generally, increasingly infiltrate the health and medical industries, a close eye should be kept on research in this area as it continues to evolve.
