Last week, while the tech titans discussed opportunities to “self-govern” their AI solutions in our nation’s capital with President Trump, 3,000 miles away the state of California enacted the first US state statute – SB 574 – known as the Court AI Protection Act, specifically regulating lawyers’ use of generative AI.
This new law takes effect on January 1, 2027 and here’s some of the key requirements:
⚖️ Lawyers may not delegate the practice of law to AI.
⚖️ If lawyers use AI, they must keep confidential and personal information out of unrestricted AI tools, verify outputs (including citations) and fix hallucinations, and disclose AI use to the court on documents they file.
⚖️ Every citation in a court filing must be personally verified by the responsible attorney.
⚖️ Arbitrators may not hand decision-making to AI and must disclose reliance on AI-generated material outside the record.
⚖️ Violations can lead to court sanctions or State Bar discipline.
It is interesting to see a new AI law which specifically applies to lawyers and it will also be interesting to see if other states (e.g., New York, Illinois) enact similar laws – a quarter of all lawyers in the US are based in California and New York
I also think that increasingly we will see more legal tech companies offer solutions to enable lawyers to operationally comply with the growing number of AI laws that are being enacted.