UT System expands Abridge ambient AI across eight health institutions

Key takeaways

• Deploy Abridge’s ambient AI platform systemwide to capture the 73% rise in recording clinicians and 80% increase in monthly notes seen in the pilot.
• Align rollout with UT REAL Health AI’s governance to maintain local decision-making while meeting responsible-AI standards.
• Use the new pre-bill review module to improve documentation integrity, coding accuracy and revenue-cycle efficiency.

Systemwide Deployment

Abridge’s partnership with the University of Texas (UT) System delivers systemwide ambient AI, turning a successful pilot into an enterprise-wide deployment across eight health institutions. The agreement builds on a pilot that grew 73% in recording clinicians and 80% in monthly notes over the last 12 months, demonstrating that scaling AI can meaningfully reduce clerical burden while preserving clinician-patient interaction.

The UT System’s eight entities—UT Southwestern Medical Center, UT Medical Branch, UT MD Anderson Cancer Center, UT Dell Medical School, UT Health Tyler, UTHealth Houston, UT Health San Antonio, and UT Health Rio Grande Valley—collectively logged nearly 11.7 million outpatient visits and more than 2 million hospital days last year. By embedding Abridge’s platform into this volume, the system creates a unified AI framework that respects the autonomy of each campus. UT REAL (Research, Engineering, Application, Lab) Health AI, launched in 2025 to advance responsible AI adoption, will oversee the governance layer, ensuring that local workflows, privacy policies, and specialty needs are addressed while the AI engine operates at scale.

Governance and Integration

Peter McCaffrey, M.D., UT REAL Health AI Chair and Vice President, framed the partnership as “a shared conviction about what healthcare delivery looks like when designed intentionally as an integrated system.” The focus is two-fold: strip away the mental clutter of clerical overhead and return that focused human presence to patients at the bedside. For administrators, the promise is clear—ambient AI can capture and structure clinician-patient conversations in real time, auto-populate documentation fields, and surface relevant clinical guidance without requiring separate data-entry steps.

From an operational standpoint, the enterprise framework pairs systemwide AI access with campus-level innovation. Abridge’s platform already supports more than 100 million patient-clinician conversations annually across 300 of the largest U.S. health systems. Extending this capacity to the UT ecosystem means each specialty—more than 50 represented in the pilot—can tap AI-generated notes, coding suggestions, and decision-support alerts within their existing electronic health record (EHR) environments. Shiv Rao, Abridge CEO and co-founder, emphasized that “true technological transformation in healthcare requires trust, workflow integration and scale,” highlighting the need for tight EHR integration and clinician buy-in.

Pre-Bill Review and Implementation Steps

The agreement also introduces Abridge’s new pre-bill review capability, designed for clinical documentation integrity (CDI), coding, and revenue-cycle teams. This module promises to flag documentation gaps before claims are submitted, potentially reducing denials and improving reimbursement rates. For revenue-cycle leaders, the tool offers a data-driven safety net that aligns clinical accuracy with financial performance.

Implementation should proceed in three ordered steps. First, each campus must finalize its local governance charter under UT REAL Health AI, defining data-use policies, model oversight, and clinician training protocols. The trade-off is the upfront time investment versus delayed launch. Second, integrate Abridge’s ambient AI APIs into the existing EHRs, starting with high-volume outpatient specialties to replicate the 73% clinician adoption surge; the risk is initial workflow disruption, mitigated by phased rollouts and real-time monitoring. Third, activate the pre-bill review module across coding and CDI teams, calibrating thresholds to balance false positives against missed errors; this adds processing overhead but can significantly cut claim rework costs.

By aligning AI deployment with responsible-AI governance, the UT System can transform documentation workflows, free clinicians for direct patient care, and strengthen revenue integrity—all while scaling a proven technology across a complex, multi-institution network.