Divya Sharma, assistant professor at the Department of Mathematics and Statistics, Faculty of Science at York University, has earned the prestigious 2025 Petro-Canada Emerging Innovator Award, recognizing her research in artificial intelligence (AI) for health care and her vision for building more equitable and inclusive systems.
The Petro-Canada Emerging Innovator Award, a joint initiative between Petro-Canada (now Suncor Energy Inc.) and York University, supports outstanding early-career faculty whose research enriches the academic environment and contributes meaningfully to society. This award celebrates excellence in teaching and research, while fostering further innovation.
Sharma is advancing research on agentic AI – a new generation of AI that goes beyond static prediction to act as a collaborative partner in decision-making. Unlike traditional systems, agentic AI assigns distinct roles to autonomous agents (e.g. physician, ethicist, patient advocate) that reason, deliberate and interact much like real clinical teams. While these systems do not inherently guarantee fairness, Sharma’s work focuses on ensuring they are designed with equity at their core, embedding safeguards so that decisions account for gender, race, language and geography.
“My research focuses on developing multi-agent AI systems that reflect how decisions are actually made in health care, through collaboration, negotiation and ethical reasoning,” says Sharma, who leads the IMPACT-AI Lab at York. “This award gives us the opportunity to build proof-of-concept systems that not only predict outcomes but also explain how and why decisions are made, with fairness at the core.”
Sharma’s team, working in close collaboration with clinicians at the University Health Network in Toronto, is already piloting applications in several high-stakes areas:
- liver transplantation, where agentic AI can simulate committee deliberations and help reduce inequities in access and outcomes;
- mental health research, where multi-agent systems could improve coordination and prioritize equity in access to care; and
- chronic disease research, where agentic AI can streamline complex genomic workflows while preserving oversight and transparency.
“As agentic AI rapidly gains attention worldwide, much remains to be done to ensure it is developed responsibly,” she adds.
Sharma’s work underscores how diverse perspectives and equity-driven innovation are essential to developing AI systems that serve all communities fairly and responsibly.
Sharma is also a principal/co-principal investigator on three recent federal grants – supported by the Social Sciences and Humanities Research Council of Canada and the Canadian Institutes for Health Research – to advance equity-focused AI in precision oncology and liver transplantation.