Optometry AI Compliance in California (2026)

Optometry practices are adopting AI-powered diagnostic tools at a rapid pace — from retinal imaging analysis to glaucoma risk screening. California's AB 489 and AB 3030 apply to these tools whenever they communicate directly with patients or generate clinical communications.

AI in Optometry: Where Compliance Applies

The optometry sector uses AI in two distinct ways: diagnostic AI embedded in clinical devices (retinal cameras, OCT scanners, visual field analyzers), and patient-facing AI in communications workflows. The compliance obligations differ significantly between these categories.

Diagnostic AI tools that assist clinicians — where the optometrist reviews results and communicates findings to the patient directly — face limited disclosure obligations under current California law. The trigger for AB 489 and AB 3030 is direct AI-to-patient communication, not AI-assisted diagnosis reviewed by a clinician.

AB 489: When Optometry AI Communicates Directly with Patients

The following optometry AI deployments require AB 489 disclosures because they involve direct AI-to-patient interaction:

  • Kiosk-based screening devices — automated vision or retinal screening stations in retail optical or primary care settings that deliver AI-generated results to patients directly
  • Teleoptometry platforms — AI intake and assessment tools in asynchronous telehealth settings where AI delivers preliminary findings before a provider reviews them
  • Patient portal chat assistants — AI bots that answer patient questions about eye health, symptoms, or test results
  • AI symptom intake forms — digital intake forms that use AI to assess vision complaints and route patients before their exam

In all these cases, the AI must disclose its non-human nature at the start of the interaction — before delivering any clinical content or assessment.

Compliance Tip

Kiosk-based retinal screening tools are increasingly deployed in pharmacies and retail optical chains. If your practice partners with a third-party screening vendor, confirm in writing that their patient-facing AI interface includes compliant AB 489 disclosures — liability can flow to the referring practice.

AB 3030: AI-Generated Patient Reports and Clinical Communications

AB 3030 applies when generative AI drafts patient-facing clinical content. In optometry, the most common triggers are:

  • AI-generated retinal health summaries sent through the patient portal
  • Automated glaucoma risk assessment reports delivered by email or text
  • AI-drafted post-exam care instructions or contact lens wear reminders with clinical guidance
  • LLM-generated explanations of diagnostic test results (OCT findings, visual field scores)

Practices must choose one of two paths: have a licensed optometrist review AI-generated content before it reaches the patient, or add a specific AB 3030 disclosure to each communication stating it was AI-generated and not reviewed by a human provider — with clear instructions for reaching the practice.

Scope Boundary: AI-Assisted Diagnosis vs. AI-to-Patient Communication

A key compliance boundary for optometry practices: an AI tool that assists a clinician in diagnosing diabetic retinopathy, but whose output is reviewed and communicated to the patient by the optometrist, does not trigger AB 489 or AB 3030. The laws target direct AI-to-patient communication. The compliance obligation arises when the AI sends information to the patient without a clinician intermediary.

Recommended Compliance Checklist for Optometry Practices

  • Audit all patient-facing AI tools: kiosks, teleoptometry platforms, patient portal bots, intake forms
  • Add AB 489 disclosures to all AI systems that communicate clinical content directly to patients
  • Map all AI-generated reports and messages delivered to patients — assign each to review or disclosure workflow
  • Review vendor agreements for screening device partners — confirm AB 489 compliance is their responsibility
  • Implement a clear pathway for patients to reach a licensed optometrist from any AI interaction
  • Document all disclosure implementations with screenshots and audit logs

Frequently Asked Questions

Does an AI retinal screening tool used in a retail optical setting require an AB 489 disclosure?

Yes. If a retail optical or optometry practice uses an AI screening device that communicates results directly to the patient — even through a printed report or digital display — and that AI system interacts with the patient directly, AB 489 requires a disclosure that the analysis was performed by an AI system, not a licensed optometrist or ophthalmologist.

Are AI-generated glaucoma risk reports sent to patients covered by AB 3030?

Yes. If a generative AI system drafts risk assessment reports, diagnostic summaries, or follow-up care instructions that are delivered to patients without a licensed provider's review, AB 3030 applies. Optometry practices must either have an optometrist review AI-generated reports before sending them, or add a disclosure stating the report was AI-generated and not reviewed by a human provider.

What disclosures are required for AI appointment scheduling and patient intake bots in optometry?

Appointment scheduling bots that only handle logistics (time selection, confirmations) without exchanging clinical information generally fall outside AB 489's scope. However, intake bots that collect symptom history, vision complaints, or health history before an exam do qualify as clinical AI interactions and require AB 489 disclosures at the start of the session.

Generate Your AB 3030 Disclosure

Use our free tool to create compliant disclosure language for your optometry practice's AI communications.

Open Disclosure Generator →