Courts across the country now face AI-generated filings that cite non-existent cases. A public database tracking these incidents has logged nearly 1,200 U.S. cases where a court sanctioned, or threatened to sanction, a litigant for improper use of generative AI.
In his latest Legal League Quarterly article, Franklin Greene, North Carolina Managing Attorney at ALAW, examines both sides of generative AI in default litigation: how courts are responding to misuse and how high-volume creditor firms can use the technology responsibly.
Where the risk lies
Greene points to In re Bryant, a recent Middle District of North Carolina bankruptcy case, where a pro se debtor’s filings included fabricated citations. The court issued a warning instead of sanctions and acknowledged AI’s potential to expand access to justice. It also made clear that every filer still has a duty to verify what they submit.
Where the value lies
For volume practices, Greene sees real efficiency gains when AI is paired with attorney oversight:
- Drafting a starting point for novel or one-off pleadings with no existing template
- Turning infrequently used documents into reusable merge templates in minutes, not hours
- Reviewing attorney-drafted work and cross-referencing legal principles
His bottom line: AI is a tool to improve human-created work, not a replacement for attorney review. Every citation gets verified. Every legal proposition gets checked.
Read the full article in Legal League Quarterly (Q3 2026) →