Agentic AI goes beyond autonomous models in healthcare. These models are beginning to function as “digital collaborators” that orchestrate tools, data and workflows to take on tasks once done exclusively by clinicians and researchers, according to a health technology expert.
Kimberly Powell, managing director of healthcare at Nvidiabelieves this shift could quickly reduce burnout and increase clinical capacity, but only once health system leaders stop thinking of AI as software and start treating it like a hireable workforce.
“Seeing these AI agents as hireable is the concept – I think it hasn’t completely clicked yet in the C-suite. They don’t see it like that – they see it as a technology at the moment. They don’t see it as, you know, employees,” she said in an interview last month.
Powell pointed out that healthcare is fundamentally workflow-driven, which makes it well-suited for agentic AI systems.
Agents are designed to mirror the step-by-step workflows that clinicians already follow, coordinating templates and tools to accomplish real work instead of just serving as passive software, she explained.
Unlike older software designed as tools for humans, agents are designed to do the work themselves. Modern software architecture has made this possible, Powell noted: She said advances like APIs and tool calls mean agents can interact with existing systems, access only the data they need, and operate within strict guardrails.
In her view, guardrails transform risky models into production-ready systems – she emphasized that AI agents are only successful by surrounding core models with domain expertise, security constraints, and up-to-date medical knowledge.
This approach reduces hallucinations and ensures regulatory compliance, giving hospital leaders confidence that AI tools can operate safely in clinical environments, Powell explained. With this assurance, adoption accelerates quickly.
While the term “agents” may still sound like a buzzword, Powell said the technology is now being used in real production, with some healthcare AI companies already expanding rapidly and offloading administrative tasks and the burdens of clinical documentation – she pointed out. Shorten And Multiply the laboratories as examples.
She expects this trend to grow even faster over the next two years.
Powell argued that agentic AI holds great potential for augmenting understaffed teams. In the coming years, she believes tool agents will increasingly be treated as employees who elevate clinicians’ practice and improve access to care.
Photo: wild pixel, Getty Images