Patient rights when AI is involved in your US healthcare

AI is now embedded across US healthcare. As of July 2025, the FDA had authorised more than 1,250 AI-enabled medical devices. Ambient AI scribes are recording clinical encounters at scale. AI is increasingly used in radiology, pathology, screening, prior authorisation, and population health management. For patients, this raises practical questions: what AI is involved in my care, what are my rights, and what do I do if AI produces a wrong result?

The legal framework

HIPAA Privacy Rule. When AI processes protected health information (PHI), HIPAA applies. You have rights to: access your medical records (including AI-influenced records); request amendments; request an accounting of disclosures; restrict certain uses; receive a notice of privacy practices that should disclose AI use. The Office for Civil Rights (OCR) enforces HIPAA, with penalties up to $1.5 million per violation category per year.

Informed consent. State medical malpractice law and professional ethics require informed consent for medical treatment. Where AI materially affects clinical decisions, disclosure may be required. California AB 489 (effective 1 January 2026) prohibits AI from implying it holds a healthcare license. California AB 3030 (effective 2025) requires baseline patient AI disclosure. Texas TRAIGA (effective 1 January 2026) includes disclosure requirements for licensed practitioners.

FDA medical device regulation. AI as Medical Device (AIaMD) and Software as Medical Device (SaMD) must meet FDA safety and effectiveness standards. Adverse events from AI medical devices can be reported through MedWatch. The FDA's January 2026 guidance clarified that AI products directly influencing clinical judgment will likely require FDA review.

Anti-discrimination. Title VI, Section 1557 ACA, and ADA apply to AI in healthcare. AI producing discriminatory outcomes triggers federal protection.

State laws. In 2025, 47 states introduced more than 250 AI bills including health AI regulation, with 33 becoming law in 21 states (Manatt Health tracking). Key 2026 themes: mental health chatbots, patient disclosure, AI presenting as clinical providers, and payor use of AI in coverage determinations including downcoding of claims.

Your specific rights

Right to know. Ask your clinician directly: "Is AI being used in decisions about my care?" While federal law does not yet mandate universal AI disclosure, state laws and professional standards increasingly expect transparency.

Right to human review. The clinician remains accountable for clinical decisions. If you have concerns about an AI recommendation, request that a human clinician review and confirm it independently.

Right to records. Under HIPAA, you have the right to access your medical records including AI-influenced clinical decisions. Providers have 30 days to comply (extendable by 30 days with notice).

Right to second opinion. A second clinician evaluating independently, without using the same AI system, provides important cross-check.

Right to file complaints. Raise concerns with: the provider's patient advocate; state medical board; HHS Office for Civil Rights (HIPAA or anti-discrimination); FDA MedWatch (AI medical device adverse events); state attorney general (consumer protection).

Payor AI, coverage and prior authorization

Health insurers increasingly use AI in coverage and prior authorisation decisions. Multiple state legislatures have addressed payor AI in 2026, including downcoding of claims. You have rights to: appeal coverage denials (ACA external review rights); request human review of AI-driven decisions; complain to state insurance regulators; pursue further appeal through state courts.

If you think AI led to a wrong result

Talk to your clinician first, ask specifically about AI involvement and the clinician's independent assessment. Get a second opinion from a clinician not relying on the same AI. Request your complete medical record under HIPAA. File a complaint through the provider's patient advocate, state medical board, or HHS OCR as appropriate. Report adverse AI medical device events via FDA MedWatch. Where AI contributed to material harm, medical malpractice and product liability remedies may be available, plaintiff-side firms are increasingly familiar with AI cases.

For mental health: California's January 2026 law bans chatbots without suicide prevention protocols. The 988 Suicide and Crisis Lifeline (call or text 988) provides direct human crisis support.

Related reading

Further reading: NIST AI RMF