Professional responsibility
California AI Guidance
California’s 2026 guidance applies existing professional duties to generative and agentic AI. Lawyers retain responsibility and independent judgment.
Practical Guidance for the Use of Generative Artificial Intelligence in the Practice of Law01
Rules 1.1, 1.3 · pp. 4–5
Keep professional judgment with the lawyer
Understand the tool’s capabilities and risks. Review, verify, and correct its output. Source
02
Rules 1.6, 1.8.2; Bus. & Prof. Code § 6068(e) · pp. 5–6
Protect client confidences
Material confidentiality risks require informed client consent. Evaluate security, access, and data use. Source
03
Rules 5.1, 5.2, 5.3; Rule 1.1 · pp. 2–3, 7
Supervise people and autonomous tools
Establish policies, training, and controls. Greater AI autonomy requires stronger supervision and verification. Source
04
Rules 1.2, 1.4 · p. 7
Communicate with the client
Evaluate disclosure in context. Follow client instructions restricting AI use. Source
05
Rule 1.5; Bus. & Prof. Code §§ 6147–6148 · p. 8
Charge for actual time
Hourly billing reflects actual work. General subscriptions typically count as overhead; disclose separately billed matter-specific costs. Source
06
Rules 3.1, 3.3 · p. 9
Review every court submission
Verify analysis and authorities. Do not permit AI to file documents or communicate with courts autonomously. Source
07
Rules 1.2.1, 8.4, 8.4.1 · pp. 6, 9
Address legal compliance and bias
Follow applicable law and address biased output, including in hiring and client screening. Source