AI Disclosure Requirements

What You'll Learn

In this module, you will learn:

Why courts are imposing AI disclosure requirements on legal filings
The landmark Mata v. Avianca case and its lasting impact on the profession
Federal court AI disclosure rules and standing orders across district courts
State-level AI disclosure requirements and ethics opinions
How to draft effective AI disclosure statements
When disclosure is legally required versus professionally recommended
Sample disclosure language you can adapt for your practice
Strategies for staying current with rapidly evolving requirements

9.1 The Catalyst: Mata v. Avianca

In June 2023, attorney Steven Schwartz submitted a brief in Mata v. Avianca, Inc. (S.D.N.Y.) containing citations to six judicial decisions. None of them existed. Schwartz had used ChatGPT for legal research and failed to verify the citations independently. Judge P. Kevin Castel sanctioned Schwartz and his colleague, imposing a $5,000 fine and requiring them to notify every judge falsely identified as authoring the fabricated opinions.

Why This Case Matters

It demonstrated the hallucination problem in undeniable terms -- fake case names, citations, quotes, and procedural histories
It established that "the AI said it was correct" is not a defense. Lawyers cannot delegate verification to a machine
It triggered a wave of judicial action. Courts nationwide began issuing standing orders addressing AI use
It put the profession on notice. After Mata, no lawyer can credibly claim ignorance of AI risks

The financial sanctions were modest, but the reputational damage was enormous. The case is now taught in law schools worldwide as a cautionary tale.

9.2 Federal Court Requirements

Following Mata, federal courts adopted AI-related rules at a rapid pace. The approaches vary, but patterns have emerged.

Common Requirements

Many federal judges now require attorneys to either certify that AI was not used in preparing filings, or disclose how AI was used and confirm that all content was verified by a human.

Notable examples:

Southern District of New York: Several judges issued standing orders requiring affirmative disclosure of AI use or certification that AI-generated content was verified
Northern District of Texas: Judge Brantley Starr issued one of the first standing orders requiring certification about AI use in all filings
Fifth Circuit: Adopted a court-wide rule requiring attorneys to certify they reviewed and verified any AI-generated content -- one of the first circuit-level rules

Practical Implications

The patchwork of requirements means you must check local rules and individual standing orders for every court where you file, monitor for updates regularly, and err on the side of disclosure when requirements are ambiguous.

9.3 State-Level Requirements

State courts and bar associations have taken varied approaches.

State court rules are emerging in jurisdictions including California, Florida, New Jersey, and Texas, each with guidance on AI disclosure expectations. State bar ethics opinions consistently emphasize that existing ethical rules apply fully to AI use, with growing attention to verification duties and confidentiality implications.

Track developments through your state bar association, the ABA's Center for Innovation, legal technology publications, and CLE programs focused on AI ethics.

9.4 Drafting AI Disclosure Statements

When disclosure is required or you choose to disclose proactively, your statement should be clear, specific, and honest.

Elements of an Effective Disclosure

Identify the specific tools used (e.g., "Claude," "CoCounsel," "Westlaw AI") rather than vague references to "AI tools"
Describe how AI was used -- research, drafting, document review, or summarization
Confirm human review by a licensed attorney
Certify accuracy of all citations, quotations, and legal assertions

Sample Disclosure Language

For court filings (when required):

The undersigned certifies that generative AI tools were used in preparing this filing, specifically [tool name] for [purpose]. All AI-generated content was reviewed and verified by the undersigned attorney. All citations were independently confirmed using [legal research platform]. The undersigned takes full responsibility for the contents of this filing.

For court filings (proactive disclosure):

The undersigned discloses that [tool name] was used to assist with [specific task]. All AI-generated content was reviewed and verified, and all citations were independently confirmed.

For engagement letters:

Our firm uses AI-assisted tools for certain aspects of legal work, including research, drafting, and document review. All AI-generated work product is reviewed by a licensed attorney before use. We maintain strict confidentiality protocols for all AI tools and are happy to discuss our practices with you.

9.5 When Disclosure Is Required vs. Recommended

Disclosure Is Required When

A court's local rules or standing orders mandate it
A specific judge's standing order requires it
A tribunal or regulatory body has adopted AI disclosure rules
Your engagement agreement includes disclosure provisions
A client specifically asks whether AI was used

Disclosure Is Recommended When

AI played a significant role in drafting substantive legal arguments
AI was used for legal research informing your strategy
Your jurisdiction has not yet adopted formal rules but is likely to
The matter involves novel legal issues where AI reliability is uncertain

Disclosure May Be Unnecessary When

Using AI-powered features built into established legal research platforms
Using spell-check, grammar tools, or basic formatting tools with AI components

When in doubt, disclose. Over-disclosure carries minimal risk. Failure to disclose when required can result in sanctions and discipline.

9.6 Staying Current

AI disclosure requirements are changing faster than almost any area of legal regulation. Strategies for staying current:

Subscribe to updates from your state bar and the courts where you practice
Review local rules and standing orders quarterly
Follow legal technology publications such as the ABA Journal, Legaltech News, and Artificial Lawyer
Attend CLE programs on AI ethics and regulation
Designate someone at your firm to monitor developments and circulate updates
Build a compliance checklist for each jurisdiction and update it regularly

Default to disclosure rather than trying to determine the minimum. Be specific. Keep records of AI use. Train your team on requirements across all jurisdictions where the firm practices.

Key Takeaways

Mata v. Avianca exposed the dangers of unverified AI-generated legal research and triggered court-imposed disclosure requirements nationwide
Federal courts have adopted a patchwork of rules through local rules, standing orders, and circuit-level requirements -- check each court where you practice
State courts and bar associations are actively adopting guidelines, ethics opinions, and disclosure mandates
Effective disclosures identify specific tools, describe how AI was used, confirm human review, and certify accuracy
Disclosure is required when mandated by courts or agreements, and recommended whenever AI plays a significant role in substantive work
When in doubt, disclose -- over-disclosure carries minimal risk, while failure to disclose can result in sanctions
Staying current requires active effort including subscriptions, CLE attendance, and designated firm monitors
Transparency is a competitive advantage -- clients and courts respect lawyers who are forthright about AI use

AI Disclosure Requirements

What You'll Learn

In this module, you will learn:

Why courts are imposing AI disclosure requirements on legal filings
The landmark Mata v. Avianca case and its lasting impact on the profession
Federal court AI disclosure rules and standing orders across district courts
State-level AI disclosure requirements and ethics opinions
How to draft effective AI disclosure statements
When disclosure is legally required versus professionally recommended
Sample disclosure language you can adapt for your practice
Strategies for staying current with rapidly evolving requirements

9.1 The Catalyst: Mata v. Avianca

Why This Case Matters

It demonstrated the hallucination problem in undeniable terms -- fake case names, citations, quotes, and procedural histories
It established that "the AI said it was correct" is not a defense. Lawyers cannot delegate verification to a machine
It triggered a wave of judicial action. Courts nationwide began issuing standing orders addressing AI use
It put the profession on notice. After Mata, no lawyer can credibly claim ignorance of AI risks

The financial sanctions were modest, but the reputational damage was enormous. The case is now taught in law schools worldwide as a cautionary tale.

9.2 Federal Court Requirements

Following Mata, federal courts adopted AI-related rules at a rapid pace. The approaches vary, but patterns have emerged.

Common Requirements

Many federal judges now require attorneys to either certify that AI was not used in preparing filings, or disclose how AI was used and confirm that all content was verified by a human.

Notable examples:

Southern District of New York: Several judges issued standing orders requiring affirmative disclosure of AI use or certification that AI-generated content was verified
Northern District of Texas: Judge Brantley Starr issued one of the first standing orders requiring certification about AI use in all filings
Fifth Circuit: Adopted a court-wide rule requiring attorneys to certify they reviewed and verified any AI-generated content -- one of the first circuit-level rules

Practical Implications

9.3 State-Level Requirements

State courts and bar associations have taken varied approaches.

Track developments through your state bar association, the ABA's Center for Innovation, legal technology publications, and CLE programs focused on AI ethics.

9.4 Drafting AI Disclosure Statements

When disclosure is required or you choose to disclose proactively, your statement should be clear, specific, and honest.

Elements of an Effective Disclosure

Identify the specific tools used (e.g., "Claude," "CoCounsel," "Westlaw AI") rather than vague references to "AI tools"
Describe how AI was used -- research, drafting, document review, or summarization
Confirm human review by a licensed attorney
Certify accuracy of all citations, quotations, and legal assertions

Sample Disclosure Language

For court filings (when required):

The undersigned certifies that generative AI tools were used in preparing this filing, specifically [tool name] for [purpose]. All AI-generated content was reviewed and verified by the undersigned attorney. All citations were independently confirmed using [legal research platform]. The undersigned takes full responsibility for the contents of this filing.

For court filings (proactive disclosure):

The undersigned discloses that [tool name] was used to assist with [specific task]. All AI-generated content was reviewed and verified, and all citations were independently confirmed.

For engagement letters:

Our firm uses AI-assisted tools for certain aspects of legal work, including research, drafting, and document review. All AI-generated work product is reviewed by a licensed attorney before use. We maintain strict confidentiality protocols for all AI tools and are happy to discuss our practices with you.

9.5 When Disclosure Is Required vs. Recommended

Disclosure Is Required When

A court's local rules or standing orders mandate it
A specific judge's standing order requires it
A tribunal or regulatory body has adopted AI disclosure rules
Your engagement agreement includes disclosure provisions
A client specifically asks whether AI was used

Disclosure Is Recommended When

AI played a significant role in drafting substantive legal arguments
AI was used for legal research informing your strategy
Your jurisdiction has not yet adopted formal rules but is likely to
The matter involves novel legal issues where AI reliability is uncertain

Disclosure May Be Unnecessary When

Using AI-powered features built into established legal research platforms
Using spell-check, grammar tools, or basic formatting tools with AI components

When in doubt, disclose. Over-disclosure carries minimal risk. Failure to disclose when required can result in sanctions and discipline.

9.6 Staying Current

AI disclosure requirements are changing faster than almost any area of legal regulation. Strategies for staying current:

Subscribe to updates from your state bar and the courts where you practice
Review local rules and standing orders quarterly
Follow legal technology publications such as the ABA Journal, Legaltech News, and Artificial Lawyer
Attend CLE programs on AI ethics and regulation
Designate someone at your firm to monitor developments and circulate updates
Build a compliance checklist for each jurisdiction and update it regularly

Default to disclosure rather than trying to determine the minimum. Be specific. Keep records of AI use. Train your team on requirements across all jurisdictions where the firm practices.

Key Takeaways

Mata v. Avianca exposed the dangers of unverified AI-generated legal research and triggered court-imposed disclosure requirements nationwide
Federal courts have adopted a patchwork of rules through local rules, standing orders, and circuit-level requirements -- check each court where you practice
State courts and bar associations are actively adopting guidelines, ethics opinions, and disclosure mandates
Effective disclosures identify specific tools, describe how AI was used, confirm human review, and certify accuracy
Disclosure is required when mandated by courts or agreements, and recommended whenever AI plays a significant role in substantive work
When in doubt, disclose -- over-disclosure carries minimal risk, while failure to disclose can result in sanctions
Staying current requires active effort including subscriptions, CLE attendance, and designated firm monitors
Transparency is a competitive advantage -- clients and courts respect lawyers who are forthright about AI use

AI Disclosure Requirements

What You'll Learn

9.1 The Catalyst: Mata v. Avianca

Why This Case Matters

9.2 Federal Court Requirements

Common Requirements

Practical Implications

9.3 State-Level Requirements

9.4 Drafting AI Disclosure Statements

Elements of an Effective Disclosure

Sample Disclosure Language

9.5 When Disclosure Is Required vs. Recommended

Disclosure Is Required When

Disclosure Is Recommended When

Disclosure May Be Unnecessary When

9.6 Staying Current

Key Takeaways

Quiz

Discussion

AI Disclosure Requirements

What You'll Learn

9.1 The Catalyst: Mata v. Avianca

Why This Case Matters

9.2 Federal Court Requirements

Common Requirements

Practical Implications

9.3 State-Level Requirements

9.4 Drafting AI Disclosure Statements

Elements of an Effective Disclosure

Sample Disclosure Language

9.5 When Disclosure Is Required vs. Recommended

Disclosure Is Required When

Disclosure Is Recommended When

Disclosure May Be Unnecessary When

9.6 Staying Current

Key Takeaways

Quiz

Discussion