South Korea's November Decision Could Make or Break the Future of Medical AI Regulation
Eleven AI health technologies face their first real-world test of clinical evidence standards, with results expected to set global precedent for balancing innovation and validation.
Illustration: AI-generated
South Korea is conducting a high-stakes experiment in medical AI regulation. In mid-November 2024, the National Evidence-based Healthcare Collaborating Agency (NECA) will announce evaluation results for 11 medical AI solutions operating under temporary approval—a decision that could either validate an innovative regulatory pathway or expose fundamental tensions between supporting AI innovation and demanding rigorous clinical proof.
The technologies under review include widely-deployed solutions like VUNO's DeepCAS, which predicts inpatient risk across 180 medical institutions and 65,000 hospital beds nationwide. Also awaiting judgment are e-AI Atrix's VitalCare and Vertis' MastCheck, a breast cancer blood test. These AI tools have been operating under South Korea's "evaluation deferral" system, which allows up to four years of non-covered service use while companies accumulate clinical evidence for formal review.
The Evidence Gauntlet
According to the original report, NECA's evaluation focuses heavily on the quantity and quality of published papers through systematic literature review, though specific thresholds remain undisclosed. VUNO, for instance, published six clinical efficacy papers during its 43-month deferral period—but whether that meets the bar remains uncertain.
The initial subcommittee review concluded September 18, 2024, with a final committee meeting scheduled for October 23, 2024. The stakes are existential: technologies that pass can proceed toward health insurance coverage, while those that fail face potential market termination.
The system has already claimed its first casualty. Genkyrix's "Genowell BCT" breast cancer test failed NECA evaluation in September 2024. The company filed a lawsuit and received a stay of execution from Seoul Administrative Court in early October, signaling that rejected companies may challenge the evidence standards themselves.
When Innovation Policy Meets Regulatory Reality
The November decisions carry implications far beyond individual companies. South Korea's Ministry of Health and Welfare has positioned AI as central to its strategy for addressing regional, essential, and public healthcare gaps—a policy vision that depends on successfully deploying validated AI technologies at scale.
But stringent evidence requirements could create a disconnect between regulatory approval standards and national policy objectives. If the evaluation deferral pathway—designed as a "valley of death" solution for AI developers—proves too difficult to navigate, it could stall government AI healthcare initiatives and undermine budget execution for the ministry's "AI Basic Healthcare Strategy."
Hospitals currently using these 11 AI solutions face uncertainty about continued access and reimbursement pathways. Negative evaluations could force discontinuation of integrated clinical workflows, disrupting care delivery while raising questions about how much real-world deployment evidence should count when formal published research remains limited.
Setting Global Precedent
This represents the first real-world test of South Korea's regulatory innovation framework for medical AI. The failure rate will establish precedent for evidence standards, potentially raising barriers for future AI developers as other nations watch how South Korea balances support for innovation against demands for clinical validation.
The outcomes will reveal whether regulatory systems designed for traditional medical devices can adapt to AI's unique characteristics—including continuous learning, deployment at scale before formal approval, and evidence generation through real-world use rather than traditional clinical trials. South Korea's November decision may ultimately answer whether innovation-friendly regulatory pathways can coexist with rigorous evidence standards, or whether that tension is fundamental and unresolvable.
Sources
Companies in this article
Reporting & accountability
How we maintain accuracy
Health AI Central separates reported facts, analysis, and public-source data. Readers can review our standards, request a correction, or consult the relevant research methodology before relying on a claim.