FOLLOW US

2. Bias and Non-Discrimination

AI may not be used to “unlawfully discriminate against or disparately impact individuals or communities” based on age, gender, race, disability, or any protected status.

3. Accuracy and Correction

Staff and judges using AI must “take reasonable steps to verify that the material is accurate” and correct any “erroneous or hallucinated output.” Any biased, offensive, or harmful content generated by AI must be removed before use.

4. Transparency and Disclosure

If a court-produced work (written, audio, or visual) is entirely generated by AI, public disclosure is mandatory via a label or watermark explaining how AI was used and which system produced the output. However, this disclosure is not required for every instance of AI assistance—only when AI creates the entire final public-facing product.

5. Compliance and Oversight

All AI use must comply with existing laws, ethical codes, and professional standards. Courts may go further, imposing stricter rules or outright bans at their discretion.

Judicial Officers and AI: Human Judgment Still Required

Standard 10.80 clarifies that for tasks “within their adjudicative role” (i.e., deciding cases, drafting opinions), judges are given strong guidance but not outright mandates. Judges “should not” enter confidential data into public AI, “should take reasonable steps” to ensure accuracy and remove bias, and “should consider” disclosure when using AI to produce public content. Notably, the standard stops short of prohibiting judges from using AI for legal research or drafting—but reminds them to uphold all ethical and statutory duties, and to never abdicate personal decision-making.

Mixed Reactions: Applause, Alarm, and National Ripples

Supporters say the rules strike a pragmatic balance—allowing courts to experiment and innovate, but requiring clear safeguards to protect privacy and prevent injustice. The Task Force emphasizes that flexibility is key: “Courts are in the best position to determine how it can meet rule 10.430’s requirements and whether its generative AI use policy should be more restrictive or detailed than the rule.”

Critics—including advocacy groups, and some judges—warn that the policy is either too permissive or too fragmented. Notable objections include:

• Transparency Fears: Some argue the rule should mandate disclosure of any AI use, not just full-AI outputs.
• Uniformity Concerns: The California Employment Lawyers Association, among others, pushed for a single statewide standard, fearing a patchwork of local rules will cause confusion and unfairness for litigants across counties.
• Job Loss & Deskilling: Unions and public commenters highlight the threat of automating critical court jobs, warning that “replacing skilled staff with automated tools…risks eroding the quality and empathy of court services.”

The Judicial Council’s response: ongoing education, phased implementation, and continued consultation with technologists, legal ethicists, and the public.

Broader Implications: Setting the National Tone

As the first U.S. court system to codify AI use, California is shaping the national debate over how far automation should go in the “people’s court.” With federal courts, other states, and legal tech companies watching closely, the state’s experience will likely influence new rules on AI disclosure, confidentiality, and fairness across America.

Some see this as a first draft—one that will require constant updates as both risks and technology evolve. As the Judicial Council notes, “Creating rules and standards to address emerging technology is challenging because those rules might become outdated, restrict innovation, or inadvertently exclude additional technologies posing the same risks the rules are meant to address.”

What Happens Next?