Federal judges started requiring AI disclosure in 2023. By 2026, the map looks completely different -- and it's still not uniform. Some districts require a standing certification. Others want case-by-case disclosure. A handful of judges have individual orders that override their district's general policy.

This is the one-page reference for every AI court disclosure requirement we track. We monitor 94 federal districts, all 13 circuits, and state courts with active AI rules. Bookmark it. Print it. Tape it next to your monitor. Because filing a brief without the required AI certification is the fastest way to start a case on the wrong foot.


Federal District Courts: The Current Count

As of April 2026, over 30 federal district courts have adopted some form of AI disclosure requirement. The approaches vary wildly. The Northern District of Texas requires a standing order certification in every case. The Eastern District of Pennsylvania uses a local rule amendment. The Northern District of Illinois requires disclosure only when AI was used for substantive legal analysis. Some districts require disclosure for any AI use including grammar checking -- which is absurd but real. The trend line is clear: more courts are adopting requirements, not fewer. If your district doesn't have one today, it probably will within 12 months.

Circuit-Level Guidance: The Appellate Layer

The Fifth Circuit was first with AI-specific guidance. The Seventh Circuit followed. Most circuits haven't issued formal AI rules but have addressed it through individual judge orders. The practical impact at the appellate level is lower because oral arguments and briefs have always required attorney certification. But the circuits that have spoken up are sending a signal to their district courts: get policies in place. If you're handling an appeal, check both the circuit rules and the specific panel judges' individual practices.

State Courts: The Wild West

State courts are even more fragmented than federal courts. Some states have statewide rules. Others leave it to individual judges. New Jersey was early with a statewide requirement. Michigan followed. But in most states, you're dealing with judge-by-judge orders that may or may not be published anywhere searchable. The safest approach for state court practice: disclose proactively in your first filing. A one-paragraph disclosure costs nothing. A sanctions order for non-disclosure costs everything.

What Courts Actually Want to See

Most disclosure requirements share three elements. First, identify whether AI was used in preparing the filing. Second, specify which AI tools were used. Third, certify that a licensed attorney reviewed and verified all citations, legal arguments, and factual assertions. Some courts go further and want to know the specific prompts used -- though that's rare and raises work product concerns. The gold standard certification is brief: 'Generative AI tools were used in drafting portions of this brief. All citations and legal arguments have been independently verified by undersigned counsel.' Adapt to your court's specific requirements.

The Penalty Landscape for Non-Compliance

Sanctions for AI non-disclosure range from embarrassing to career-threatening. Mata v. Avianca set the tone with $5,000 sanctions in SDNY. The Portland case hit $109,000 in sanctions. But the real penalty isn't financial -- it's reputational. Judges remember. Opposing counsel remembers. The legal media definitely remembers. Every AI-related sanctions case gets national coverage. Compare that to the 30 seconds it takes to add a disclosure paragraph. The math is not complicated.

The Bottom Line: AI court disclosure is expanding, not contracting. The firms that build disclosure into their filing workflow now -- as a default, not an afterthought -- won't have to scramble when their next court adopts a requirement. Check your specific jurisdiction using our individual court pages linked throughout this guide.

AI-Assisted Research. This piece was researched and written with AI assistance, reviewed and edited by Manu Ayala. For deeper takes and the perspective behind the research, follow me on LinkedIn or email me directly.