California’s New Legal Rules for 2026: AI, Photo Proof of Service, and Simpler Statements of Decision

New statutes and court rules taking effect in 2026 and 2027 will change how California lawyers serve papers, preserve appellate issues, and disclose their use of artificial intelligence. Appellate attorneys Tim Kowal and Jeff Lewis focus on what actually matters in practice—what to fix now, and where the new traps are likely to appear.

The big changes:

• AI in the Courts: Rule of Court 10.430 requires courts to either ban AI use by judicial officers and research attorneys or adopt a formal AI policy with verification and disclosure requirements. Expect cautious policies, broad disclosures, and little tolerance for “the AI did it” excuses.
• One Deadline for Statements of Decision: AB 515 eliminates the short-trial/long-trial distinction. If you want a statement of decision, you must request it before submitting…and you should do it in writing.

Other changes worth noting:

• Photo Proof of Service: Starting January 2027, AB 747 requires process servers to document service attempts with photographs showing GPS coordinates and timestamps.
• Court Reporter Disclosure: AB 711 requires meet-and-confer declarations to disclose whether court reporter attendance was discussed and the outcome.
• Electronic Service Authorized: SB 85 allows courts to approve service by email or electronic means when traditional service fails.
• Expanded Mediation Authority: Courts may order mediation in cases up to $75,000 if at least one party requests it and no discovery disputes are pending.
• AI Disclosure in Bankruptcy Court: The Southern District of California Bankruptcy Court now requires disclosure of AI tools used and certification of independent accuracy review.

Listen now to understand what to change in your templates and where the next procedural missteps are waiting.

AI-sanctions might get eyeballs, but the bigger sanctions are still for plain old bad lawyering. Jeff also raises this ethical and pragmatic question: who defends the lawyer when sanctions threaten the client? Should counsel facing an OSC retain separate counsel for the sanctions component to avoid divided attention and better protect client interests? What if the costs of independent counsel are likely to exceed the sanction?

• $25K for using Anti-SLAPP as a litigation weapon. A bare-bones anti-SLAPP was amplified by record emails suggesting the strategy was to inflict cost and pain rather than win on the merits.
• $13K for relitigating the merits through a fee appeal. The appeal purported to challenge fees, but largely recycled arguments already rejected in the prior appeal. The court finds the effort both objectively meritless and subjectively aimed at rehashing settled ground.
• <$2K for fabricated authority & thin explanations. Schlichter v. Kennedy results in $1,750 against an attorney who relied on nonexistent or inapposite citations and offered credibility-challenged explanations about verification methods. After the court’s exhaustive point-by-point teardown, the monetary sanction seems merciful.
• Pro per’s sanction is dismissal of appeal. In Arno Kuglua v. Young Park, the Court of Appeal dismisses an appeal for failure to support arguments with proper authority.

Also: AI guidance from the courts**:** The California Courts of Appeal publish user-facing AI guidance emphasizing verification, independent source-checking, and personal accountability for filings—even if AI assists with drafting.

A scandalous Netflix documentary called an unconventional sex-based therapy business an “orgasm cult,” all based on a sole source whose account has several flaws. But the Court of Appeal dismissed the defamation case on anti-SLAPP grounds. Tim and Jeff discuss whether any California defamation case against a media company could survive the one-two punch of anti-SLAPP and NY Times v. Sullivan. They also discuss California’s unique approach to standing—it’s not jurisdictional, it’s purely pragmatic.

• Anti-SLAPP meets documentary defamation: OneTaste Inc. v. Netflix illustrates how courts evaluate actual malice when the plaintiff is treated as at least quasi-public, and how journalistic discretion can sink a claim even where the plaintiff says it provided contrary evidence before publication. Tim flags the built-in squeeze: if public-figure status and the controversy are intertwined, the plaintiff may need discovery to prove merit, but cannot get discovery without first showing merit.
• Standing without injury, by design, not accident: Kashanian v. National Enterprise Systems tees up a standing fight over technical FDCPA disclosure issues, think small-font compliance, with no alleged real-world harm. The takeaway is not subtle: in California, legislative authorization can do a lot of work, and no harm does not necessarily mean no case.
• When the statute creates the bounty, sanctions become the guardrail: The hosts debate whether CCP 128.5 and CCP 128.7 actually deter nuisance filings when the underlying enforcement scheme invites penalty-driven litigation. Is it appropriate—or wise—to use our courts as civil bounty enforcement, devoid of any harm requirement?
• Juror privacy is real, ask the team that wrote the $10,000 check: Don’t research prospective jurors on social media.
• Minute entry, real consequences: A timing skirmish over whether a minute entry can function as an appeal-triggering order ends, for now, with the U.S. Supreme Court declining review. Be conservative in calculating the time to appeal