The Centers for Medicare & Medicaid Services’ (CMS) Health Tech Pledge marks a significant industry milestone. This pledge represents commitment by leading healthcare and tech organizations to make patient data accessible through standardized, interoperable systems. If successfully implemented, providers, pharmacies, and health systems will be able to access unprecedented volumes of data through Fast Healthcare Interoperability Resources (FHIR) by July 2026.
However, simply having greater access to patient data won't automatically improve patient outcomes. Without new methods to parse and summarize electronic health records (EHRs), the very interoperability intended to enhance patient outcomes could worsen issues such as alert fatigue, information overload, and data quality problems.
To help ensure the success of the CMS Health Tech Pledge in improving patient outcomes, healthcare organizations should consider pairing EHR connectivity with artificial intelligence (AI) summarization tools. These tools will help transform the overwhelming amount of data into practical, useful insights.
Interoperability Challenges and the Promise of AI
Interoperability and data standardization address the technical challenge of information sharing, but they do not solve the issue of clinical useability, especially in the heat of a medical emergency.
Many doctors know the difficulty of parsing actionable insights from an EHR — an issue exacerbated by the scale of patient data at both the individual and institutional level. A single record can contain decades of lab results, images, and notes, and the average hospital produces around 50 petabytes of data per year according to a 2019 World Economic Forum study. That number has almost certainly grown in the intervening years. For context, 50 petabytes is the equivalent of 50 million average-sized books. The entire Library of Congress, by comparison, only contains about 17 million books.
AI tools will play an important role in extracting actionable information from this sea of patient data.
In a clinical setting, different stakeholders often need to access different information from within the same dataset. An emergency room physician might require a patient’s immediate risk factors, while a primary care provider would need comprehensive health information and preventive care reminders. Patients themselves might need a completely different summary: a plain-language overview of their diagnosis and action items for follow-up care.
Context-aware AI summarization tools will be able to deliver situation-specific summaries using the same underlying models. Imagine a patient comes into the emergency room with a broken leg. An effective AI assistant would surface this patient’s history of recurring blood clots over the past six years — information that would change the treatment approach and medication load. More important, however, is the information the AI tool didn't show: an alert about the patient’s lapsed flu vaccine status, which would be irrelevant in this emergency.
Effective tools will also need to extend beyond simple summarization. They will be able filter out duplicative information, identify and exclude outdated data, prioritize outputs based on clinical context, and present takeaways in language appropriate for the recipient. At scale, AI summarization tools can bridge the gap between CMS-driven interoperability and frontline usability, enabling health systems to convert standardized data flows into tangible improvements to patient outcomes, care coordination, and clinical efficiency.
The Imperative of Accuracy and Oversight
Before AI tools receive broadscale use throughout the healthcare industry, they must meet rigorous accuracy standards. The hallucinations, omissions, misedits, and output quality issues that plague many AI models can pose significant dangers in healthcare settings. Adding an extra word or omitting key details can lead to critical medical mistakes while increasing the fatigue, stress, and workloads of practitioners as they seek to validate output accuracy.
AI tools are only as good as their underlying data. Low-quality data compounds the risk of AI errors. Summarization tools must be able to identify and resolve poor recordkeeping, misspellings, copy paste errors, acronyms, and more. The sheer size of EHRs adds further complexity. A patient’s health record typically contains hundreds of thousands of tokens worth of data, which can far exceed many models’ context windows and require massive computing power to parse.
For these reasons, AI tools can augment but not replace professional human judgment. Oversight and clinician review remain essential. Continuous monitoring and feedback loops, regular model updates and training, and clear audit trails for AI-generated summaries will be critical to enforcing quality control. Implemented together, these safeguards create what safety professionals often call a “Swiss Cheese Layer” of protection. No single layer is flawless, but overlapping defenses reduce the chances that an error will pass through uncorrected.
Integrating AI into the care continuum will take time, trial, and concerted evolution on the part of providers, administrators, doctors, and patients. But the advantages are clear: AI provides a path to improve care delivery in a new age of EHR interoperability and data standardization.
Preparing for 2026 and Beyond
The CMS Health Tech Pledge is not the first push for interoperability in healthcare, but it represents a major and promising initiative designed to solve the issue of fragmented and inaccessible healthcare data. By July 2026, organizations have committed to providing data access through FHIR APIs, returning key patient data from chart notes, clinical documents, and history. The date is not far away. Providers need to start thinking now about how they’re going to turn that data into clinically useful information.
Healthcare organizations that excel in AI summarization will thrive in the new data-rich environment. Those that fail to adapt risk being overwhelmed by information overload. The time to act is now. Healthcare leaders must start planning for AI-enabled workflows today to seize the competitive and medical advantages that the CMS initiative will create tomorrow.