Earning Trust for AI in Health 2025

Healthcare tech companies are working to accelerate the development of high-quality AI technologies that meet the needs of the health sector across the world. Clear regulatory frameworks and supporting guidelines will be critical to foster the purposeful innovation of health AI technologies. Unlocking this new approach will require three strategic shifts: 1. Build technical expertise among health leaders and clinical decision-makers Health leaders and clinicians should seek to upskill and engage with technical experts with healthy scepticism, actively challenging technical propositions to ensure that they align with the overarching vision. In the future, understanding the capabilities, limitations and risks of AI technologies will no longer be the sole responsibility of chief technology officers (CTOs) but will become a fundamental skill for health leaders and clinicians to adapt evaluation practices to the presence of AI technologies.34 2. Support the translation of legislative goals into actionable guidelines that create incentives for purposeful innovation The emergence of the first generation of AI- focused legislation establishes a paradigm within which the use of AI technologies is considered acceptable. The next step of developing complementary guidance documents and infrastructure can benefit significantly from public–private engagement, such as the organization of regulatory sandboxes, rigorous evaluation methods including pre- and post-market surveillance and AI assurance resources to detect early signals of AI-related risks as soon as possible and with full transparency. Trust can be earned even before legislation comes into effect by adhering to existing guidelines and standards. 3. Mobilize public–private partnerships to actively engage the private sector in lifecycle management Private-sector involvement in AI systems’ evaluation efforts is important due to the rapidly evolving AI innovation landscape. PPPs are necessary to engage with the private sector in order to cope with the increasing number of AI technologies that need to be tested and must be compliant with a growing set of requirements. In addition, these partnerships can play a crucial role in supporting the acceleration of model training and development as well as post-deployment monitoring. Promoting cooperative engagement such as public–private partnerships and prioritizing upskilling and evaluation practices can create an innovation environment that is agile and transparent. Collaborative action can build a system that not only harnesses AI to revolutionize healthcare but does so in a way that prioritizes patient safety and trust. The future of AI in health has immense promise, and with collective effort, society can ensure that it delivers on that promise responsibly.Conclusion Health-system leaders, regulatory bodies and the private sector must collaborate to unlock AI’s full potential while mitigating its associated risks.