Song Dow Lee v. HSBC: C.D. California Threatens $2,500 Sanction for AI-Misrepresented Cases + Failure to Correct

Background: A Banking Dispute and Citation Misrepresentations

Song Dow Lee and co-plaintiffs sued HSBC Bank USA relating to a banking/wire transfer dispute, apparently challenging the bank’s pre-contractual representations and its duty to provide advertised services. When defendants moved to dismiss, plaintiffs filed an Opposition that defendants alleged in reply contained multiple case misrepresentations and a non-existent quotation.

Plaintiffs subsequently filed a “Notice of Errata” intended to address the issues — but instead introduced additional inaccuracies. Then plaintiffs filed a Proposed Statement of Decision on April 15, 2026 without correcting the identified misrepresentations. Judge Walter independently reviewed the sources and found defendants’ objections warranted.

The AI Issue: Failure to Correct After Notice

The most significant AI dimension of this case is the post-notice failure to correct. After defendants identified misrepresentations in their Reply, plaintiffs failed to correct them in the subsequent Proposed Statement of Decision. This mirrors the Amtrust v. Liberty Mutual pattern: once a party has been put on notice of AI citation errors, continuing to file without correction is an independent Rule 11 violation. The court found “several cases were cited for unsupported propositions; some quoted language could not be located in the cited sources; and other quotations contained inaccurate pincites.”

Judge Walter’s show cause order required counsel to address, in their response, “whether artificial intelligence was used to draft Plaintiffs’ Opposition and Proposed Statement of Decision, and what actions counsel took to determine that any of the propositions of law relied on by Plaintiffs were accurate.” This mandatory AI disclosure in a show cause response is becoming a standard judicial inquiry.

Holdings

1. Show cause order issued — $2,500 sanction or dismissal. Counsel must show cause in writing why the court should not impose $2,500 in sanctions and/or dismiss the action.
2. AI disclosure required in show cause response. The court specifically required the response to address whether AI was used and what verification steps were taken.
3. Post-notice failure to correct is the aggravating factor. The court found that plaintiffs’ failure to correct acknowledged misrepresentations in their subsequent Proposed Statement of Decision was the triggering event for the show cause order.

“Plaintiffs’ counsel are ordered to show cause… why this Court should not impose sanctions on counsel in the amount of $2,500.00 and/or dismiss this action… [The response] shall address, inter alia, whether artificial intelligence was used to draft Plaintiffs’ Opposition and Proposed Statement of Decision.”

— Judge John F. Walter, C.D. California, May 7, 2026

India Angle: Post-Notice Failure to Correct in Indian Banking Litigation

India’s banking and financial services litigation — before the Banking Ombudsman, Debt Recovery Tribunals, NCLT for insolvency matters, and civil courts — involves complex factual and legal submissions where AI assistance is tempting. The Song Dow Lee pattern — failure to correct identified citation errors in subsequent filings — is particularly relevant because Indian courts keep track of what has been raised and corrected across proceedings.

Relevant Indian Law

• Banking Regulation Act 1949 / RBI Guidelines: Banking cases frequently turn on precise statutory provisions and regulatory circulars. AI tools routinely misstate or hallucinate RBI guideline references, particularly FEMA and wire transfer provisions under FEMA 1999 and its amendments.
• BCI Rule 22: An advocate who knows of an error in a submission must correct it at the earliest opportunity. Failure to correct after the opposing party points out an AI citation error — as occurred in Song Dow Lee — would be a breach of Rule 22 in an Indian context.

Three Practical Tips

1. When opposing counsel identifies citation errors, correct them immediately. Filing a subsequent document without correcting identified errors — as the Song Dow Lee plaintiffs did in their Proposed Statement of Decision — is the act courts find most sanctionable. In India, a similar failure to correct would breach BCI Rule 22 and could be treated as contemptuous persistence in error.
2. A “Notice of Errata” that introduces new errors is worse than no correction at all. The Song Dow Lee plaintiffs’ errata introduced additional inaccuracies. Any correction you file must be comprehensively verified — do not file a correction without checking all citations in the corrected document, not just the ones originally flagged.
3. AI disclosure is now standard in show cause responses in U.S. courts — Indian practitioners should pre-disclose. Courts are now routinely requiring AI disclosure as part of show cause responses. Indian practitioners can get ahead of this by proactively noting any AI tool used in brief preparation in their covering letter or at the foot of the submission, together with a statement that all citations have been independently verified.

Quick Takeaways

• Post-notice failure to correct AI citation errors — continuing to use impugned citations in subsequent filings after defendants identified them — is the most serious aggravating factor in this case.
• A $2,500 sanction or dismissal of the action is the threatened consequence — courts are treating persistent post-notice AI errors as equivalent to deliberate misrepresentation.
• Courts are now routinely requiring AI disclosure in show cause responses — practitioners should expect this as a standard element of any AI hallucination enforcement proceeding.
• In India: failure to correct identified errors in subsequent submissions may breach BCI Rule 22 and could be used as evidence of contumacious conduct in contempt proceedings.