Protecting Your Firm from the Use of AI-Generated Fake Citations
Artificial intelligence has become a valuable tool in modern legal practice. It can assist with research, drafting, summarization, and administrative tasks. However, as many courts and law firms have recently learned, AI also carries significant risks—particularly when it generates false or fabricated legal authorities, commonly referred to as “AI hallucinations.”
The legal profession is witnessing an increasing number of cases involving attorneys who unknowingly submit briefs containing non-existent cases, fabricated quotations, or mischaracterized authorities generated by AI platforms. Every law firm should take proactive steps to protect itself from the professional, financial, and reputational consequences that may arise when an attorney relies on AI-generated citations without proper verification.
The Growing Problem of AI Hallucinations
In April 2026, Sullivan & Cromwell, one of the nation’s premier law firms, publicly apologized to a federal judge after submitting a filing containing numerous errors generated by artificial intelligence, including fabricated case citations and inaccurate legal authorities.
Similarly, the Ninth Circuit Court of Appeals recently suspended two attorneys for six months after they submitted court filings containing AI-generated hallucinations.
California courts have also begun imposing sanctions where attorneys have relied upon fabricated legal authorities. In Noland v. Land of the Free, L.P. (2025) 114 Cal.App.5th 426, 430, the Court of Appeal found an appeal frivolous where:
“Nearly all of the legal quotations in plaintiff’s opening brief, and many of the quotations in plaintiff’s reply brief, are fabricated. That is, the quotes plaintiff attributes to published cases do not appear in those cases or anywhere else. Further, many of the cases plaintiff cites do not discuss the topics for which they are cited, and a few of the cases do not exist at all.”
The message from courts is clear: attorneys remain responsible for the accuracy of every citation and legal authority submittedto a tribunal, regardless of whether AI assisted in preparing the document.
Every Firm Should Adopt a Written AI Use Policy
The most effective way to reduce risk is to establish clear written policies governing the use of artificial intelligence.
A comprehensive AI policy should include the following provisions:
1. Require Approval Before Using AI Tools
Employees should obtain approval from firm management before using generative AI platforms for firm business.
Only approved AI platforms should be used, and all AI accounts should be created using firm-authorized credentials.
Employees seeking to use additional AI tools should submit a written request explaining how the technology will be used.
2. Require Paid Enterprise Versions
Whenever possible, firms should require the use of paid or enterprise-grade AI products rather than free public versions.
Enterprise products generally offer stronger privacy protections and reduce the likelihood that user data will be incorporated into future model training.
3. Mandatory Verification of All AI-Generated Content
No attorney or staff member should rely upon AI-generated legal content without independent verification.
If a citation, legal proposition, statute, regulation, or factual statement cannot be independently confirmed through reliable sources such as Westlaw, Lexis, official court opinions, or government publications, it should not be used.
4. Prohibit Entry of Confidential Client Information
Public AI platforms should never receive confidential client information, litigation strategy, privileged communications, or sensitive firm data.
This safeguard protects client confidentiality and helps ensure compliance with Rule 1.6 and ABA Formal Opinion 512.
Acceptable Uses of AI
When properly supervised, AI can be a valuable productivity tool.
Within Approved Legal Research Platforms
Attorneys may use AI-enabled legal research products such as Westlaw Precision AI for:
Drafting internal memoranda and templates.
Summarizing legal concepts, rules, and case law.
Organizing research findings.
Brainstorming legal theories and arguments.
All citations and authorities must still be independently verified.
Within Approved ChatGPT Enterprise Platforms
Provided no confidential client information is entered, attorneys may use AI to:
Draft internal memoranda.
Prepare non-confidential marketing content.
Generate policy drafts.
Assist with administrative tasks.
Brainstorm operational improvements.
A useful rule of thumb is:
Treat AI like a summer law clerk—helpful, efficient, and capable of generating useful work product, but never fully responsible for the final result.
Prohibited Uses of AI
Law firms should expressly prohibit:
Entering client information into public AI platforms.
Submitting AI-generated legal arguments without independent review.
Citing cases generated by AI without verification through Westlaw, Lexis, or another authoritative source.
Using AI as a substitute for legal analysis and professional judgment.
Misleading courts with false or unverified information.
Using AI tools without informing the supervising attorney when required by firm policy.
Required Safeguards
Every attorney and staff member should follow these safeguards:
Use only firm-approved AI platforms.
Ensure all AI-generated content undergoes human review before being sent to clients or courts.
Verify every citation, statute, quotation, and legal proposition.
Submit all briefs and legal research through Westlaw Cite Check or similar citation verification tools.
Document significant AI use in internal matter notes when appropriate.
Seek supervisory guidance whenever uncertainty exists.
As with most professional responsibility issues, asking questions is far safer than making assumptions.
Consequences of Misusing AI
Improper use of AI may result in:
Mandatory retraining.
Internal disciplinary measures.
Court-imposed sanctions.
Termination of employment for reckless or repeated violations.
Conclusion
Artificial intelligence is not going away. Properly used, it can improve efficiency and reduce costs. Improperly used, it can expose attorneys and law firms to sanctions, malpractice claims, disciplinary proceedings, and reputational harm.
The solution is not to ban AI. The solution is to govern it responsibly.
Every law firm should adopt clear written policies, require verification of all AI-generated citations, train attorneys regarding their ethical obligations, and maintain robust supervisory procedures. By doing so, firms can enjoy the benefits of AI while minimizing the risks associated with hallucinated legal authorities and fabricated citations.
The attorney—not the AI program—remains responsible for every word filed with a court.
