Artificial intelligence (AI) models used in healthcare may recommend different treatments for the same medical condition based solely on a patient’s socioeconomic and demographic characteristics, according to a new study.
Researchers created nearly three dozen fictional patient profiles and tested nine large language AI models across 1,000 simulated emergency room scenarios. Although the clinical details remained the same, the AI models occasionally altered their recommendations based on personal factors like income and ethnicity.
The findings, published in Nature Medicine, revealed disparities in care recommendations. Doctors more often recommended advanced diagnostics like CT scans or MRIs for high-income patients, while they routinely discouraged low-income patients from undergoing further testing—reflecting real-world healthcare inequalities. The researchers observed these issues across both proprietary and open-source AI models.
“AI has the power to revolutionize healthcare, but only if it’s developed and used responsibly,” said Dr. Girish Nadkarni of the Icahn School of Medicine at Mount Sinai. Coauthor Dr. Eyal Klang added, “By identifying where these models may introduce bias, we can work to refine their design, strengthen oversight, and ensure patients remain at the heart of safe, effective care.”
New Hope for Treating Sjogren’s Syndrome Symptoms
In a breakthrough, scientists have advanced toward easing the severe dryness caused by Sjogren’s syndrome, an autoimmune disorder affecting the eyes and mouth.
According to Reuters, researchers studying mice discovered that the protein tricellulin—responsible for binding salivary and tear gland cells—breaks down early in the disease’s progression. As a result, this cellular damage reduces saliva and tear production.
Two experimental treatments—an investigational drug (AT1001) and a novel molecule—successfully restored saliva production in mice. One therapy repaired the damaged cell junctions, while the other prevented the breakdown from starting. Both approaches revived normal gland function.
“This changes how we think about treating Sjogren’s syndrome,” said lead author Dr. Xin Cong of Peking University. “We’re moving beyond simply calming inflammation. Now we can fix the actual structural damage in the glands. What’s even more encouraging is that both approaches worked, which gives us real confidence in developing patient-ready therapies.”
Experimental Drug Tolebrutinib Slows MS Progression
Researchers have identified a promising treatment for non-relapsing secondary progressive multiple sclerosis (SPMS)—a form of MS with no currently approved therapies.
Tolebrutinib, an oral Bruton’s tyrosine kinase (BTK) inhibitor developed by Sanofi, delayed disability progression by 31% in a large clinical trial involving 1,131 patients. At six months, 22.6% of patients receiving tolebrutinib showed confirmed disability progression, compared to 30.7% in the placebo group.
Additionally, more patients in the tolebrutinib group experienced improvements in disability (8.6% vs. 4.5%), and the drug also reduced markers of inflammation and tissue damage.
However, the drug poses safety concerns. Liver complications were more frequent among users, with about one in 200 patients developing severe liver enzyme elevations in the first three months of treatment.
Carefully monitor patients,” warned study lead Dr. Robert Fox of the Cleveland Clinic. “Stop the drug immediately if liver enzymes rise.
In two other trials involving patients with relapsing MS, tolebrutinib did not outperform Sanofi’s existing treatment, Aubagio (teriflunomide), in reducing relapse rates.