Georgia Experts Warn: Unregulated Medical AI Could Exacerbate Health Disparities

ATLANTA, Ga. (Atlanta News First) — With artificial intelligence increasingly influencing healthcare, certain specialists warn that without proper regulation, it might exacerbate current inequalities, especially affecting minority groups.

Georgia is currently grappling with a significant lack of healthcare providers. The Cicero Institute reports that out of the state’s 159 counties, 143 are dealing with staffing deficiencies, and the demand is projected to require around 8,000 more physicians by 2030.

Although AI is not anticipated to take over for physicians, it is being brought in to alleviate stress on overloaded practices by managing paperwork and aiding in quicker diagnostic processes.

Several AI tools are currently being utilized. For example, AI programs assist in examining mammography images to assess the risk of breast cancer or help surgeons make decisions during operations in real time.

“A great deal of potential lies with AI,” stated Jessica Roberts, who teaches law at Emory University. “However, we might recreate the same issues that arise from our human decision-making processes.”

Roberts cautioned that the healthcare information utilized to develop AI systems isn’t consistently inclusive. A significant portion of this data originates from white doctors examining white patients, which implies that these technologies might yield skewed or limited outcomes when applied to people of color.

She worries about how it impacts escalating issues such as the maternal mortality crisis.

"When examining the Latino community and African Americans, they constitute approximately 35% of the U.S. population, yet represent just around 7% of healthcare providers," stated Dr. Cecil Bennett.

African American women in Georgia succumb to complications related to pregnancy at almost triple the rate of Caucasian women, and research indicates that their discomfort is frequently overlooked by healthcare providers, resulting in delayed treatment.

"There can be increased challenges at times in communication between doctors and patients when their cultural backgrounds differ," stated Bennett.

Roberts and other supporters are urging for more stringent rules to guarantee that AI tools are trained using varied data sets and created considering fairness.

"We must ensure that when utilizing AI and incorporating it into our systems, we do so in an equitable manner," Roberts stated.

Roberts feels that developers have an obligation to ensure their algorithms accurately represent the communities they aim to serve.