The State Bar of California has proposed changes to six Rules of Professional Conduct that would create new, explicit expectations for lawyers who use AI in their work. While these rules are not yet final, the direction is unmistakable. Firms that wait for formal adoption to adjust their practices will already be behind. This article explains what California is proposing, how these changes might work in day to day practice, and why trademark clearance work is squarely in scope.
What California Is Proposing
California's ethics regulators are not simply issuing high level guidance about AI. They have drafted amendments to multiple rules of professional conduct that would:
- Require attorneys to independently verify AI outputs before relying on them in client matters;
- Require disclosure to clients when AI plays a material role in the work being done for them; and
- Require firms to adopt internal AI governance procedures, including supervision and policies.
These are not aspirational best practices. They are proposed rules of conduct, which means that once adopted, ignoring them may result in disciplinary action.
The timing is deliberate. Courts are already penalizing lawyers for AI related mistakes, and California is signaling that it does not intend to wait for a wave of mishaps before putting standards in place.
When AI Mistakes Become Ethics Problems
Recent cases show how quickly AI errors can have real consequences for lawyers and their firms. In one widely discussed example, a U.S. magistrate judge sanctioned the managing partner of a firm after a junior lawyer submitted a brief that included a fake case citation generated in part by an AI research tool. The partner did not draft or sign the brief, but his name appeared on the filing. The court imposed a monetary sanction and mandatory training, and it made one point unmistakably clear: supervising attorneys remain responsible when AI related mistakes reach the court.
California's proposed amendments essentially encode that expectation. If AI contributes to your work product, you own the result, and you will be expected to show that you exercised judgment rather than mindlessly accepting what the tool produced.
What This Means for Firms
The risk is not limited to disciplinary exposure. It also touches client trust, institutional reputation, and even how pitches and RFPs will be evaluated.
Many firms already use AI in research, drafting, due diligence, and contract review. The proposed rules do not prohibit those uses. Instead, they demand that someone be clearly accountable for the output and that the firm have a defensible process around AI use.
In practice, that means firms must be able to answer questions like:
- Which AI enabled tools are being used across the firm, and by whom?
- What verification steps are required before AI influenced work is sent to a client, filed with a court, or submitted to an agency?
- How is AI use being documented and, when appropriate, disclosed to clients?
Firms that cannot answer these questions consistently are behind on compliance and at risk of uncomfortable conversations with courts, regulators, and clients who are now attuned to AI related risk.
Why Trademark Clearance Is Directly in Scope
Trademark clearance is a textbook example of work where AI is already making a measurable difference and where the new rules will matter.
Clearance work is high volume and high stakes. AI tools can search multiple databases, identify similar marks, and flag possible conflicts much faster than any human reviewing records one by one. They can group visually or phonetically similar marks, score apparent risk levels, and surface patterns that might otherwise be missed.
The danger is not the speed. It is the temptation to treat AI results as final answers instead of as starting points for legal judgment. Under California's proposed rules, an AI generated clearance report would almost certainly trigger duties of verification and, often, disclosure. The attorney who signs off on a clearance opinion would be expected to:
- Review the underlying records and reasoning, not just accept a list of "safe" or "risky" marks;
- Confirm that obviously relevant marks have not been overlooked or misclassified; and
- Decide whether the AI's role in the search and analysis is significant enough that the client should be told about it.
In other words, the way the AI tool is designed and deployed becomes just as important as how fast it returns results.
How CrossBeamIP Approaches AI and Ethics
CrossBeamIP was built with these concerns in mind. The platform ingests trademark data daily from the USPTO, the Madrid Protocol, and other key sources, then uses a custom-built AI engine to generate practical clearance insights. It scans candidate marks, groups similar results, and flags potential conflicts across different mark types. The goal is to let attorneys see, at a glance, where the risks are most likely to be.
Just as important, the system maintains a clear audit trail. For each flagged result, attorneys can see why it was surfaced, how it relates to the client's mark, and how it fits into the broader analysis. That makes it straightforward to review and verify each finding before anything is shared with a client.
This is a very different experience from a barebones USPTO search. A standard image search for a logo might return hundreds of loosely related records with no prioritization, no risk scoring, and no explanation. A paralegal or junior lawyer then must manually triage everything, often under time pressure.
By contrast, CrossBeamIP starts with a guided client questionnaire, so the AI runs the search in the right commercial and strategic context. The platform then generates a professional, downloadable, well-organized report designed for attorney review and client communication. The workflow is AI assisted analysis with attorney review built in; it is not AI replacing the attorney's judgment. That is exactly the pattern California's proposed rules are pushing firms toward.
The Takeaway for Trademark Practices
California's proposal is a preview of where professional responsibility is headed across the United States. Firms that build verification, disclosure, and governance into their AI workflows now will be seen as responsible and trustworthy. Firms that rely on opaque, unreviewed AI outputs will face tougher questions from clients, courts, and state bars.
For trademark practitioners, the message is simple: AI can and should make clearance work faster and more insightful, but only if the underlying process makes it easy for attorneys to verify results, explain how the analysis was done, and stand behind the final opinion.
The rules are not final yet, but the direction of travel is clear. Now is the time to bring your AI-enabled trademark clearance process in line with where ethics and client expectations are headed, rather than waiting for the next disciplinary headline to force the issue.