The healthcare industry has long awaited the transformative power of technology, but so far, the outcomes have been disappointing, with increased administrative tasks and complexity. However, Jaimon Jose, senior director of engineering and compliance at CharmHealth, believes the current situation is distinct. The Menlo Ventures 2025 State of AI in Healthcare report reveals that healthcare providers are now investing over $1 billion annually in AI adoption, marking a significant shift towards large-scale deployments. This trend is driven by the emergence of agentic AI, which enables autonomous systems to reason, plan, and act, potentially transforming the industry’s approach to enterprise AI adoption.
Connecting the Dots Across the Ecosystem
Unlike earlier AI tools that simply asked users to take action, agentic AI systems now handle entire workflows independently. Healthcare processes are rarely straightforward, often requiring coordination between patient intake, record-keeping, testing, billing, and follow-up—each with its own separate systems. By integrating across these fragmented steps, agentic AI creates smoother, more reliable patient experiences. The real opportunity in agentic AI isn’t tied to any single company’s product but to the underlying systems that allow different tools to work together effectively.
A major obstacle to progress in healthcare has long been the isolation of data across electronic health records, billing platforms, scheduling tools, and external databases. Initiatives like the model context protocol (MCP) are breaking down these barriers by enabling AI to link external resources seamlessly. According to McKinsey’s research, open frameworks such as MCP let AI agents tap into operational data across healthcare networks.
In practical terms, MCP allows AI to retrieve patient files, verify medication interactions, book appointments, or process claims—tasks that were previously cumbersome to automate. Within a year of its November 2024 launch, MCP surged from about 100,000 downloads to over 8 million by mid-2025. Last December, it was transferred to the newly established Agentic AI Foundation under the Linux Foundation.
Thanks to MCP’s standardized approach for running interactive interfaces, AI tools now deliver tailored data visualizations, decision workflows, and documentation aids directly within care settings. These include real-time patient trend tracking, guided documentation forms, and automated approval workflows, all dynamically adjusted to each situation. While these innovations are embedding AI into daily operations, human judgment remains essential. No algorithm can fully replicate clinical expertise, ethical responsibility, or compassion. The goal of agentic AI is to assist, not replace, providers by reducing mental and administrative workloads, creating a collaborative dynamic among tech developers, healthcare teams, and patients.
This system ensures technology providers build secure, transparent, and verifiable tools, while healthcare organizations carefully adopt them into their operations. Clinicians must retain final authority over diagnoses, treatments, and care plans, with every medical decision requiring human oversight. Patients also play a critical role: the effectiveness of AI-driven care depends on their willingness to provide thorough and accurate health details, including family history, prescriptions, allergies, and symptoms. When patients engage actively with their care teams, the AI working behind the scenes can offer more precise, personalized support.
The Financial and Human Impact of Adoption
The convergence of agentic AI, open interoperability standards, and healthcare’s pressing need for efficiency creates a significant opportunity. Industry analysts predict that agentic AI in healthcare will grow 40-45% annually. This expansion moves beyond simple software upgrades into a fundamental reshaping of operational capacity. The focus now shifts to how organizations implement these systems to handle the volume of work required.
