AI Governance Tops ECRI’s 2025 List of Healthcare Hazards—A Call for Vigilance
Click to Enlarge
The ECRI Institute’s annual report on health technology hazards highlights the urgent need for healthcare organizations to adopt robust AI governance in 2025. The report warns that over-reliance on AI-enabled health technologies, without adequate scrutiny, could lead to inappropriate care decisions. While AI offers transformative potential in diagnostics and efficiency, challenges such as algorithm bias, hallucinations, and model drift demand ongoing performance monitoring.
ECRI emphasizes that only well-funded institutions, such as large academic medical centers, typically have the resources to implement comprehensive AI oversight, leaving smaller and rural facilities at risk of falling behind. To mitigate these risks, ECRI calls for healthcare organizations to define AI goals, establish governance protocols, and ensure data readiness for algorithm use.
Other critical hazards include unmet technical support for complex medical devices used in homecare settings, which is ranked second on the list. Improper device setup and insufficient user training could jeopardize patient outcomes in hospital-at-home or chronic condition management programs.
Cybersecurity threats, especially those involving third-party vendors, rank third. ECRI points to the February 2024 cyberattack on Change Healthcare, which disrupted claims processing across the industry, as a stark reminder of the vulnerability of relying on external vendors for essential operations. The institute urges healthcare organizations to vet third-party vendors and rigorously assess security risks.
Healthcare leaders are encouraged to prioritize these safety concerns and establish protocols that ensure patient safety, system resilience, and responsible adoption of new technologies.
Healthcare industry needs to get a grip on AI governance in 2025, ECRI warns (Beavins, FierceHealthcare, 12/4).
