Duty of Candor (AI Citations)

Model Rule 3.3 prohibits knowingly making a false statement of fact or law to a tribunal and requires correcting material false statements previously made. Submitting fabricated citations violates it, and the obligation to correct attaches once the lawyer learns of the error.

Courts have treated unverified AI output as a failure of the lawyer's own diligence rather than a technology malfunction.

Alternative Names:

Candor and AI Output, AI Citation Candor

Why it Matters?

This is the obligation that has produced actual sanctions, and repeatedly. The pattern is consistent: a lawyer uses a general-purpose model for research, does not read the cited cases, files the brief, and cannot produce the authorities when challenged. The failure is verification, not AI, and courts have shown no appetite for the distinction.

Frequently Confused with

Related terms

Frequently asked questions

Is submitting a fabricated citation an ethics violation even if unintentional?

Courts have found that failing to verify before filing breaches the lawyer's diligence and candor obligations. Lack of intent affects the severity of the response, not whether a violation occurred.

What must a lawyer do upon discovering a filed citation was fabricated?

Correct it with the tribunal promptly. Rule 3.3 imposes an affirmative duty to remediate material false statements, and delay compounds the exposure substantially.