Shomique Completes His UNIQ+ Project with AI4DH

On 22nd August, 2026, Shomique Hayat completed his UNIQ+ research project with the AI4DH Lab on multi-agent AI for clinical decision-making.

Shomique's project investigated whether discussion between AI agents can improve antibiotic recommendations for patients with bloodstream infections. Using real-world intensive care data and laboratory antimicrobial susceptibility results, his work showed that disagreement between AI agents could substantially change recommendations even without new clinical evidence, sometimes leading to greater use of reserve antibiotics without improving antimicrobial coverage.

Shomique presented his work at the UNIQ+ Presentation Event, sharing his findings with students and researchers across the University of Oxford. His project highlights the importance of evaluating how clinical AI systems respond not only to evidence, but also to interaction and disagreement between AI agents. We thank Shomique for his excellent contribution during the summer and wish him every success in his future studies and research.