Medical Reform in the AI Era
A U.S. agenda assuming AI reaches doctor-level or better performance across many tasks.
-
Medical Freedom
Protect autonomy: Informed consent, treatment refusal, second opinions, lawful choice.
Enable direct AI access: Independent AI diagnosis, test interpretation, treatment comparisons, and health planning without mandatory physician gatekeeping.
Reform prescriptions: Base restrictions on actual harm, monitoring, and misuse risks—not precedent.
Expand treatment access: Simplify off-label, investigational, and Right to Try pathways, especially for seriously ill patients lacking alternatives.
Protect self-experimentation: Reconsider low-risk testing and interventions while preventing fraud, contamination, and third-party harm.
-
AI Infrastructure
Approve by outcomes: Test real-world performance across demographics and conditions.
Enable AI prescribing: Authorize validated scopes; escalate high-risk or uncertain cases.
Continuously evaluate: Monitor outcomes, update safely, disclose performance changes, address failures; build on FDA frameworks.
Ensure data portability and control: Interoperable records, meaningful consent, strong privacy, limits on commercial reuse.
Clarify liability: Distinguish developer, clinician, institutional, and patient responsibility.
-
Lower Costs
Publish all-in prices: Cash, negotiated, facility, and out-of-pocket costs.
Restore competition: Scrutinize consolidation, exclusionary contracts, patent abuse, and barriers to cheaper alternatives.
Reform PBMs: Disclose fees and rebates; curb incentives favoring expensive drugs; pass savings to patients. The FTC has flagged competition concerns.
Accelerate generics and biosimilars: Remove unnecessary barriers, address unjustified evergreening, reduce approval delays.
Automate routine care: AI triage, common conditions, test interpretation, and monitoring.
Enable direct purchasing: Compare providers, labs, pharmacies, and tests without unnecessary intermediaries.
Reform reimbursement: Pay for validated outcomes and remote care, not unnecessary visits.
-
Open Innovation
Enable personalization: Modernize individualized therapy, small-batch production, and compounding rules without sacrificing quality.
Modernize trials: AI-assisted design, adaptive trials, appropriate synthetic controls, real-world evidence.
Open research: Expand access to de-identified data, publications, validated methods, and reproducible research.
Enable independent replication: Test health claims without manufacturer dependence.
Create regulatory sandboxes: Time-limited testing of low-risk AI and care models with transparent oversight.
Compounding reform must enable genuine personalization without exposing patients to poor-quality or inadequately tested products.
-
Risk-Based Safeguards
Regulate proportionately: Differentiate wellness advice, routine tests, antibiotics, and invasive experiments.
Strengthen high-risk protections: Major surgery, toxic treatments, irreversible interventions.
Preserve informed consent: Explain uncertainty, evidence, alternatives, interactions, and harms.
Maintain product quality: Prevent contaminated, counterfeit, or deceptively labeled drugs and devices.
Prevent coercion: Protect against data exploitation and pressure from employers, insurers, and institutions.
Measure outcomes: Track adverse events, accuracy, treatment success, satisfaction, and prices—not just compliance or visits.
Five Priorities
Universal AI access: Affordable diagnosis, explanations, second opinions.
Prescription reform: Direct access and AI prescribing where supervision adds little safety.
Competition and transparency: Comparable prices, fewer barriers, direct purchasing.
Open innovation: Faster trials, personalized treatments, and research with proportionate quality standards.
Autonomy and safeguards: Protect choice and privacy; focus oversight on serious, irreversible, or imposed risks.
Principle
Regulate risk, not the desire to manage one's own health.
AI should not require historical human intermediaries. But diagnostic superiority alone cannot establish treatment safety: quality, interactions, contraindications, and long-term outcomes still matter.
Goal: Nearly free medical knowledge, automated routine care, risk-based drug restrictions, bodily autonomy, and high-cost care competing on measurable value.
AI enables reform; systemic change delivers affordable, innovative, free medicine.