Artificial intelligence (AI) stands at the forefront of a potential revolution in health care, promising to significantly enhance patient outcomes, bolster safety, and make high-quality services more accessible and affordable worldwide. Yet, despite these promising prospects, the journey toward fully integrating AI into health systems faces substantial hurdles. Achieving this future depends on overcoming challenges related to data quality, infrastructure, trust, and incentivization—factors that are critical for realizing AI’s true capabilities across diverse populations and health settings.

A collaborative effort by the Future of Health (FOH), an international network of senior health leaders, and the Duke-Margolis Institute for Health Policy has shed light on the key areas requiring urgent focus. Their comprehensive review and consensus-building activities highlight four primary action areas: enhancing the standardization and quality of health data, developing robust infrastructure for safe and effective AI development and evaluation, promoting data sharing while safeguarding privacy, and aligning financial incentives to accelerate AI adoption. These recommendations aim to guide health systems and policymakers worldwide in unlocking AI’s potential to revolutionize health care delivery.

The Critical Role of High-Quality Data in AI Development

The foundation of effective AI in healthcare is reliable, comprehensive data. As one FOH member emphasized, “Data will become the most valuable commodity in health care,” underscoring the importance of strategic data management. The performance of AI algorithms is directly linked to the accuracy, completeness, and representativeness of the data they are trained on. To support this, stakeholders must prioritize identifying and ensuring access to key data elements that hold the highest predictive value and clinical significance, such as patient outcomes and safety metrics.

Efforts like the multinational STANDING Together initiative exemplify the importance of standardizing data collection to improve AI reliability and fairness. Policy measures, including aligned payment incentives, can further motivate health organizations to focus on high-quality data collection, emphasizing the importance of trustworthy and timely information for AI applications. Such measures will be vital for building a solid foundation for AI-driven innovations that can truly improve health system performance.

Ensuring Trustworthiness and Ongoing Verification

The effectiveness of AI tools hinges on their safety and performance in real-world settings. As highlighted by FOH members, “Your algorithms’ reliability depends on the quality of data and the policies that uphold privacy and trust.” AI’s performance can vary significantly across different sites and over time due to changes in data patterns—a phenomenon known as data drift. For example, studies of the Epic Sepsis Model have demonstrated performance disparities based on location and time, emphasizing the need for continuous monitoring and recalibration.

Developing consensus-based evaluation procedures and establishing validation mechanisms are crucial steps. Initiatives like HealthAI, which aims to certify AI validation processes, and the planned US-wide network of health AI assurance labs, are promising developments. These efforts will help ensure that AI tools meet safety standards and comply with emerging regulations such as the EU AI Act and the US Executive Order on AI. Building infrastructure for ongoing performance assessment and bias detection is essential to maintain trust and efficacy.

Interesting:

Facilitating Data Sharing Across Boundaries

The power of AI is amplified when data can be shared seamlessly across institutions and borders. As one expert noted, “Standardizing business processes is fundamental to making data usable across different healthcare systems.” Interoperable, longitudinal data enables AI models to capture diverse patient experiences, improving accuracy and generalizability.

Countries like South Africa are making advances with initiatives such as the National Digital Health Strategy, which aims to create portable health records and standardized data flows. Promoting privacy-preserving sharing methods—such as federated learning, synthetic data, and dynamic consent—can foster trust among patients and providers. Engaging communities and ensuring patients have access to their own health data further enhances transparency and participation, vital for sustainable AI integration. Establishing clear governance frameworks and trusted third-party data stewardship can expedite data access, enabling rapid innovation while protecting individual privacy.

Aligning Incentives to Accelerate AI Adoption

One of the most significant barriers to AI progress in health care is the lack of direct financial incentives. As one leader pointed out, “Without compensation, organizations are less motivated to invest in high-quality data and AI tools.” Shifting toward value-based care models, which reward outcomes rather than services, offers a pathway to align financial incentives with AI-driven improvements.

Implementing payment structures that reward quality, safety, and efficiency can catalyze the development and deployment of AI solutions. High-quality, standardized, and interoperable datasets are fundamental to measuring and enhancing care quality, especially when integrated into value-based payment schemes. Such models motivate health organizations to prioritize trustworthy data collection, innovative AI development, and continuous performance improvement, ultimately leading to better patient outcomes and a more sustainable health system.

Moving Toward an AI-Enabled Future in Healthcare

The trajectory of AI in health care depends heavily on the strategic actions of leaders and policymakers. Early successes demonstrate AI’s potential in transforming specific clinical practices, yet realizing broader benefits requires concerted efforts to improve data infrastructure, foster trust, and realign financial incentives. By focusing on these key areas—standardized data, rigorous evaluation, secure sharing, and appropriate reimbursement—health systems can move from incremental progress to a true AI-driven revolution.

The future of health care hinges on the ability to harness AI responsibly and effectively, ensuring that technological advances translate into tangible health benefits for populations worldwide. As the landscape evolves, continuous collaboration among stakeholders and adherence to best practices will be essential to navigate the complexities and unlock AI’s full potential.

Additional Resources