Northwestern Brings Generative AI to the Bedside

The health system’s 2025 move to embed AI in its EHR signals a turning point for clinical adoption and oversight debates

24 Feb 2026

Artificial intelligence is edging out of the lab and into the exam room. In September 2025, Northwestern Medicine announced a partnership with Tempus AI to embed a generative AI tool called David directly into its electronic health record system, marking a notable shift from experimentation to everyday use.

Unlike earlier tools that lived on separate platforms or flooded doctors with pop up alerts, David works inside the patient record itself. Clinicians can generate short summaries, highlight key medical issues, and ease documentation tasks without toggling between screens. The goal is simple: cut paperwork and surface useful insights in real time.

For Northwestern, the move is as much about people as it is about technology. Health systems nationwide face staffing shortages and mounting administrative demands that sap time and morale. By folding AI into daily workflows, leaders hope physicians can spend more time with patients and less time wrestling with forms.

Tempus AI has been careful in how it frames the tool. The company says generative AI is meant to support doctors, not replace them, a message that resonates in a field wary of automation outpacing judgment. Trust remains the currency of healthcare, and no algorithm can substitute for clinical experience.

The broader industry is watching closely. While many hospitals use AI for imaging and predictive analytics, far fewer have embedded generative tools at the core of clinical operations. If Northwestern can show measurable gains in efficiency and quality, competitors are likely to follow.

Still, the road ahead is not frictionless. Health leaders must ensure AI generated content is accurate, secure, and transparent. Regulators are paying closer attention, and questions about cost, oversight, and liability will shape how quickly similar systems spread.

Even so, the direction is clear. Generative AI is becoming woven into the fabric of care delivery, and its success will hinge less on novelty than on responsible use. Northwestern’s bet suggests the future of medicine may be written, at least in part, by code working quietly behind the scenes.