How AI and innovations in payment models will impact healthcare in 2025  

Updated on January 23, 2025
Businessman connecting tech devices and icons applications with each other 3D rendering

Payment model innovations and the ongoing deployment of artificial intelligence (AI) technologies will be among the key trends in healthcare in 2025 as more providers and payers embrace value-based care (VBC) programs designed to produce better clinical outcomes while lowering costs of delivery. 

In addition, life sciences companies, payers, device manufacturers, academic researchers, and public health organizations can be expected to put into production various flavors of AI – such as predictive AI, generative AI, machine learning (ML), and natural language processing (NLP) – to drive advances in VBC programs, care management, clinical use cases, drug development, resource allocation, supply chain management, predictive analytics and operational efficiency. 

Below are five key areas we can expect to have a major impact on healthcare in 2025:  

Debut of the ACO PC Flex Model 

The voluntary ACO Primary Care (PC) Flex Model was created by the Centers for Medicare and Medicaid Services (CMS) and launched on New Year’s Day. PC Flex is designed to enable a more equitable, innovative, and team-based approach to care by steering more healthcare dollars toward underserved populations, such as rural Americans. To reduce health inequities, it is essential that primary care practices have the flexible funding they require to identify and address people’s unmet health-related social needs, implement strategies to remove barriers to high-quality primary care, and improve care coordination.  

Applying AI and machine learning to digitized data will enable healthcare organizations to build collaborative care networks that produce better clinical results, reduce costs, and boost health equity. We can expect to see more of these initiatives from healthcare organizations in 2025. However, a robust and scalable cloud-based data architecture is necessary to support a healthcare hierarchical “network of networks.” Such an infrastructure would enable payments and the easy exchange of secured patient data to increase care coordination and efficiency.     

Continued adoption of VBC payment models 

VBC adoption was on the upswing in 2024, and 2025 should see more of this trend as healthcare organizations upgrade their abilities to extract value from the massive amounts of digital health data – most of which is unstructured – found in electronic health records (EHRs).  

Digitized and tagged properly across different data ontologies, health data can fuel the development of models that link care delivery to outcome-based reimbursement, which would accelerate VBC adoption. This can be especially useful in creating targeted individualized outcome-based care plans encoded into Clinical Quality Language (CQL) to help with reporting metrics and payment requirements. 

Payment model innovation  

Without tight alignment of primary and specialty care, the goals of VBC – better outcomes and lower costs of care – are virtually unattainable. Given that specialty care represents roughly 60% of total care costs in the U.S., emphasis in the new year will be on models that encourage patient-centered care and address care coordination. 

As one example, episodes of care that focus on a single condition will be expanded upon by CMS with the Transforming Episode Accountability Model (TEAM) – a mandatory, episode-based alternative payment model from the CMS Innovation Center – and also within the commercial and direct-to-employer space. The goal of these measures is to reduce redundant services and improve patient outcomes along with the patient experience.  

Under more global programs, population health initiatives that emphasize patient outreach and navigation will enable prospective management of high-needs and polychronic patients who drive the largest share of costs. Further, benefit designs will be evaluated for alignment with value-based programs to ensure consistency of approach and effective incentives for both patients and providers in 2025 and beyond.  

Harnessing AI and machine learning 

Gains in computing power coupled with the emergence of advanced AI algorithm models now provide for deep insights – both retrospectively as well as prospectively – that can inform care decisions. Key AI technologies that healthcare organizations will continue to deploy in 2025 include:  

  • 1)  Speed, efficiency, and reduction in errors on the tasks performed by humans: Automating claims handling, preauthorization and appeals, patient onboarding, scheduling, and automated documentation 
  • 2)  Advances in drug discovery, drug efficacy and side effects, and medical research: AI will start to bridge the gap of speed, scale, and tools for execution 
  • 3)  Advanced home monitoring and wearable devices usage in care coordination activities: AI modules can analyze the data, produce summary charts/readings for caregivers, make recommendations on diet and exercise regimens/adjustments, alert clinicians and patients on medication reminders/changes/prescription refills, etc. 
  • 4) Gen AI adoption with CQL: Helps with the development of care models to link care delivery to outcome-based reimbursement models, thus helping speed up VBC adoption; this can really help with the creation of targeted individualized, outcome-based care plans, which can be encoded into CQL to help with reporting metrics and payment requirements 
  • 5)  AI will assist in patient safety monitoring systems: Different sensing technologies like computer vision, pressure sensors, vital-sign monitoring and wearable devices will start to improve patient safety across different adverse events that cause patients harm. 

Zero Trust framework adoption 

The Change Healthcare ransomware attacks in 2024 that crippled the healthcare industry highlighted the weaknesses and vulnerabilities of many infrastructures, platforms, and solutions. As healthcare generates and stores more patient health data and protected health information (PHI), there will be a need for additional investments in cloud infrastructure and digital tools. To ensure health data is protected, healthcare organizations will need to implement robust cybersecurity processes and systems.  

A good way to identify and implement controls is via annual certifications of SOC 2 Type II, HITRUST, FedRAMP, CMMC and NIST. While these are critical aspects of good cyber hygiene, a robust and routine training and awareness program for employees is essential. 

For platforms and solutions, the implementation of Zero Trust Frameworks will continue to see massive adoption. Zero Trust works on the principle that no user, device, or application should be trusted by default, even if they are within the network perimeter. 

A key first step in applying this model is the implementation of micro-segmentation to ensure that interactions between entities are highly secured by isolating different parts of the network. Continuous network traffic monitoring and anomaly detection are important to identify and prevent potential breaches, unauthorized access attempts, malware infections, and other suspicious activities. 

Controlling access to data and managing authentication, authorization, encryption, and least-privilege access controls are critical aspects of information security. Multi-factor authentication (MFA) is a pivotal tool in achieving Zero Trust Security. MFA requires users to submit two or more forms of authentication that fall under these four categories: knowledge (PIN), inherence (biometrics like fingerprint, voice, etc.), device possession (USB key, token, etc.) and location (via GPS tracking). With these safeguards in place, internal and external security penetration tests by third parties will ensure robustness of the infrastructure, platforms and solutions.  

One other area that firms would start utilizing even more is to implement machine learning as a key strategy to combat phishing attacks, leveraging its capacity to analyze, adapt, and recognize patterns that may signify harmful activities. This technology can monitor user behavior, analyze the content of emails to identify potential phishing signs, and flag spam or malicious messages. 

ML also can examine URL domains and links to highlight suspicious ones, providing users with prompt feedback about the safety of links or attachments, thus improving their security awareness. Moreover, it can be incorporated into current security frameworks, including firewalls and intrusion detection systems, to establish a comprehensive defense strategy against phishing attacks.

Rahul Sharma headshot copy
Rahul Sharma
CEO at HSBlox

Rahul Sharma is chief executive officer of HSBlox, which assists healthcare stakeholders at the intersection of value-based care and precision health with a secure, information-rich approach to event-based, patient-centric digital healthcare processes – empowering whole health in traditional care settings, the home and in the community.

Lynn Carroll 20221110 111308 copy
Lynn Carroll
Chief Operating Officer at 

Lynn Carroll is the chief operating officer of HSBlox, which assists healthcare stakeholders at the intersection of value-based care and precision health with a secure, information-rich approach to event-based, patient-centric digital healthcare processes – empowering whole health in traditional care settings, the home and in the community.