California May Make AI Ethics Enforceable: Inside the Proposed Rule Amendments (and How to Prepare)
9/20/20262 min read
California May Make AI Ethics Enforceable: Inside the Proposed Rule Amendments
Most states that have addressed AI in legal practice have done it through ethics opinions — useful, but advisory. California is on a path that could look different. A set of proposed amendments to the California Rules of Professional Conduct would, if adopted, place specific AI obligations into enforceable rule text carrying disciplinary authority.
Nothing here is final. The proposals are pending. But the direction is worth understanding now, because the gap between "recommended practice" and "sanctionable violation" is exactly what these amendments would change.
Two tracks, one easy point of confusion
California is running two parallel efforts, and it's easy to blur them:
The Practical Guidance (updated and approved in May 2026) is advisory. It tells you how existing duties apply. The proposed rule amendments are the second track. COPRAC advanced proposed changes to several Rules of Professional Conduct at its March 13, 2026 meeting and opened a public comment period that closed in early May 2026. If the process runs its course and the California Supreme Court adopts them, these would become binding rules — not suggestions.
That's the headline: California is considering moving AI duties from guidance into the disciplinary rulebook. As of now, that step has not been taken.
What the proposals would do
Based on the public comment materials, the amendments would, among other things:
Add the use of artificial intelligence to the internal policies that managing lawyers must establish — placing AI governance alongside longstanding obligations like conflict screening, calendaring, and safeguarding client funds. Require lawyers to verify the accuracy and existence of the authorities they cite before filing, expressly including authorities generated or assisted by AI. Clarify that the instruction and supervision lawyers owe to nonlawyer personnel extends to AI use in the delivery of legal services.
Read together, the proposals would convert several themes from the advisory guidance into duties a lawyer could be disciplined for ignoring.
Why this matters more than another ethics opinion
An advisory opinion signals how regulators think. An enforceable rule sets a floor with consequences. If adopted, California firms would need documented AI governance not because it is prudent, but because a managing lawyer's failure to have it could itself be a rule violation. The verification duty would give a hallucinated citation a clearer disciplinary hook than "general" competence and candor obligations alone.
How to prepare without over-committing
Because the amendments are pending, the goal is readiness, not overhaul:
Treat a written AI governance policy as table stakes now, so adoption doesn't catch you flat-footed. Build a standing pre-filing verification step into your litigation workflow — it's already good practice and would map directly onto the proposed duty. Extend your supervision structure to cover how staff and nonlawyers use AI, in writing. Assign someone to watch the docket on this: track the State Bar's rulemaking so you learn the moment anything is finalized.
None of these steps depend on adoption. All of them reduce risk today and position you to comply immediately if the rules take effect.
FAQ
Are these AI rules in effect in California? No. They are proposed amendments and remain pending; they are not currently binding.
How would they differ from the Practical Guidance? The guidance is advisory. These amendments, if adopted, would carry disciplinary authority.
When will California decide? No adoption date is settled. The proposals went through public comment in 2026 and remain in the rulemaking process.
We help California firms build AI governance that satisfies today's guidance and is ready for tomorrow's rules. If you want a policy that won't need a rewrite the day these amendments are adopted, let's talk.