Value today comes from efficiency improvements and reduced friction across both patient and provider experiences. Rather than replace clinical judgment, leading GenAI use cases automate routine tasks, streamline clinical interactions and make it easier for patients to get the information and guidance they need. Common applications fall into three broader categories: provider support, patient support and operational efficiency. 

While these use cases create value across the care journey, the potential of GenAI in health care depends on moving from fragmented applications to a cohesive, interoperable ecosystem that connects data, workflows and decisions across the care journey. This can multiply the impact of AI, enabling personalized care, smarter resource allocation and more thorough care decision support.  

Provider support:  

• Imaging analysis: GenAI can help accelerate the review of medical images and X-rays, which can support faster identification of potential abnormalities, earlier diagnoses and reduced wait times for patients. 
• Ambient listening: GenAI can listen to patient-provider conversations and automatically generate notes for each visit. This enables providers to spend less time typing and focus more on listening and engaging. This can improve the interaction on both sides, helping patients feel heard and reducing providers’ documentation fatigue.  
• Automated coding: GenAI-powered ambient listening also facilitates accurate and automated coding. While listening to patient-provider conversations, GenAI tools can convert visit details into billable codes, reducing manual effort and minimizing incomplete or delayed documentation for care support.  
• Patient communication support: Providers have also found value in using GenAI to support patient communication. Tools can help generate clear, compassionate explanations of diagnoses, treatment options or next steps based on the patient’s situation and level of understanding. This can make complex information easier to absorb and help patients feel informed and reassured.   

Patient support:  

• Diagnostics and triage: GenAI tools have advanced past simple symptom checkers. Patients can use these tools to gauge symptoms, evaluate certain risks and understand when they need to escalate to a physician. Modern GenAI tools can ask follow-up questions and gather medical history to create a SOAP (subjective, objective, assessment and plan) note—a structured method to document patient encounters—that patients can take to their provider to enable continuity of care. These tools can also help guide patients toward next steps, such as booking a specialist appointment.  
• Care navigation and decision support: Once patients have a diagnosis, they often need help determining and moving forward with next steps. Alongside a patient's care team, GenAI can help fill that gap by guiding patients through what to do next, whether that includes finding a specialist, understanding treatment options or deciding whether their issue is suited for a primary care appointment or urgent care visit. This support can help the health care journey feel less fragmented and more manageable for patients.  

Operational efficiency: 

• Contact center support: GenAI can help answer common patient questions and support appointment scheduling. This allows contact center teams to handle higher call volumes more efficiently, freeing human agents to focus on more complex or sensitive interactions.  
• Fraud detection: Health care organizations see GenAI’s potential to reduce fraud by moving beyond post-payment recovery to near real-time fraud prevention. GenAI can help identify abnormal patterns (e.g., unusual usage of services or benefits) so organizations can intervene sooner and reduce downstream losses.   
• Administrative support: Routine administrative tasks, such as requesting insurance details, checking claim details, updating contact information or confirming appointments, can take up valuable staff time. GenAI-powered tools can streamline these interactions, reducing staff time spent on low-value tasks and helping organizations proactively correct inaccuracies that could cause issues later.  
• Scheduling: GenAI can also support workforce and scheduling processes, including checking shift coverage and tracking paid time off (PTO). This can create meaningful efficiency gains and help practices run more smoothly. 

Patient care inherently requires trust and accuracy, so some common obstacles to AI adoption in health care stem from concerns around liability and reputational damage. Organizations want to feel confident that the tools will protect patients while delivering measurable value. Obstacles can include: 

• Regulatory and liability constraints: Many health care organizations are uncomfortable using GenAI to provide clinical recommendations, as they are concerned about liability if an AI-generated output is inaccurate or leads to a poor outcome.  
• Clinician risk aversion: Even though GenAI shows potential, some clinicians may hesitate to use it, especially in clinical workflows, because they don’t want to risk reputational damage or hinder patient trust. This has generally resulted in faster adoption in administrative areas and slower adoption in clinical workflows. 
• Data readiness: Achieving value from GenAI depends on strong underlying data. However, many organizations lack clean, structured, centralized data. Preparing data for scale often requires significant upfront investment, which can give health care organizations pause—especially if they’re not confident in how they will demonstrate the return on investment (ROI) of GenAI tools.   
• Difficulty measuring ROI: Many health care organizations struggle to measure GenAI value in ways that support sustained investment. Benefits are typically not immediate, and measuring impact often involves collecting data or results over multiple years, by which time patients may have left the system. In addition, capturing ROI often requires significant operational changes. Without that broader transformation, value can be difficult to demonstrate.

Health care organizations that effectively implement and scale GenAI now can create value for their business and patients. Balancing the adoption of new technology with the intricacies of patient care requires a cautious and disciplined approach, which should include:

• Establishing strong guardrails: Health care organizations should be intentional about where they want GenAI to operate. For many use cases, the goal is to enhance care navigation and support decision-making without delivering direct diagnoses or treatment recommendations. Organizations should build strong guardrails around these goals by designing safe responses that avoid offering medical advice and route clinical questions to physicians.   
• Maintaining human-in-the-loop oversight: Human review remains essential in health care workflows. As organizations scale GenAI, they should continue to incorporate human-in-the-loop oversight. For example, physicians should continue to carefully review and confirm accuracy of GenAI-generated visit summaries before sending them to patients. GenAI tools can streamline processes and save time, but providers must stay involved in the process to review, validate and make final judgments.   
• Redesigning experiences: Health care GenAI implementations generate meaningful results when organizations redesign experiences around the technology, rather than layering AI onto legacy processes. Treating GenAI as an add-on to existing workflows often limits impact. Leaders can create greater value by rethinking processes end to end and using GenAI to remove friction on a foundational level.  
• Starting with targeted use cases and then scaling: Given the risks involved and clinician hesitancy, organizations can benefit from first implementing GenAI in targeted ways that have clearly defined boundaries and measurable KPIs. Starting small gives teams the opportunity to learn in a contained environment, build confidence and demonstrate the value of GenAI before expanding the technology to other areas and applications. Organizations can use these first implementations to understand their governance needs and risk tolerance, so they can scale effectively over time.

We’re seeing exciting pilots and implementations across the health care industry that combine back-office efficiency with patient journey improvements. For example:

• A major health system transformed its billing and collection calls with conversational AI. Now, 15%+ of those contact center’s 4M+ annual calls are fully automated. The updated virtual agents saw 16% growth in dollars collected compared to the previous version of the agents. 
• A leading health insurer is reimagining high-volume grievances and appeals (G&A) process review with AI. Its AI-powered reasoning engine gathers and normalizes inputs and provides criteria-backed clinical summaries that providers can review and validate. This is designed to speed review cycles and reduce manual effort while maintaining human oversight.