Twenty-Eight Percent
Johns Hopkins tested every leading AI agent on real clinical work. Only 28 percent of complex tasks finished on the first attempt.
Johns Hopkins ran frontier AI agents against real clinical workflows last week. Documentation chains, operational tasks, the work staff do every day. Twenty-eight percent of complex tasks completed on a first attempt. Fewer than 8 percent of agents stayed consistent across repeated runs.
Johns Hopkins changed the scope. Teams that broke multi-step processes into single, bounded tasks — one agent, one handoff, one check — got completion rates above 90 percent. The improvement came from narrowing the ask.
That math matters. An agent completing a simple task at 94 percent reliability across hundreds of daily cases produces more than one attempting a twelve-step process and failing seven times out of ten. Reliable completion compounds across the workload.
This tracks with a broader pattern. Ninety-seven percent of executives have deployed agents. Twenty-nine percent report ROI. The gap is scope. Most teams went from vendor demo to production and mistook deployment for results.
The fix is asking a smaller question. One that the agent can actually answer, every time, at production quality. Then asking the next one.