4 MIN READ
Published 09/17/2026
Until recently, settlement was a sign of closure, and dismissal with prejudice meant a claim was finished. But Nippon Life Insurance of America v. OpenAI suggests those assumptions are no longer as durable as carriers have long thought them to be — not because the law changed, but because the cost of generating litigation collapsed.
In March, Nippon Life Insurance Company of America sued OpenAI in the U.S. District Court for the Northern District of Illinois, bringing claims for tortious interference with contract, abuse of process, and the unlicensed practice of law. While courts spend the next several years testing novel legal theories, the underlying fact pattern should grab the attention of every liability carrier now.
How a closed claim reopened
Claimant Graciela Dela Torre had settled a long-term disability dispute with Nippon Life in January 2024. She signed a release, took the settlement payment, and her case was dismissed with prejudice. Months later she grew dissatisfied. Rather than return to counsel, she uploaded her attorney’s correspondence to ChatGPT and asked whether she was being manipulated. The tool told her yes, according to the complaint. She fired her lawyers and used the chatbot to draft a motion to reopen the settled case under Rule 60(b).
When the court denied that motion, Dela Torre filed a new lawsuit against other parties, then amended it to include Nippon again and reassert the same released claims. By Nippon’s count, dozens of motions, subpoenas, and requests for judicial notice — many of which, the complaint alleges, served no legitimate purpose — flooded in. At least one relied on a court decision that does not exist, a fabricated citation produced by AI and filed by the litigant. Nippon says their response has cost roughly $300,000. The big story lies in the numbers, not the legal novelty.
Cost disparity creates real exposure
For most of the history of civil litigation, producing a motion carried a natural governor: It took a professional’s time and that cost money. This friction kept filing volume roughly proportional to the stakes and merits of a dispute. Generative AI tools, however, remove the governor. A dissatisfied party can now generate a professional-looking motion, brief, or discovery demand in minutes at almost no marginal cost.
The defense has no such relief. Every filing on a docket still has to be read, evaluated, and — in most cases — answered. Silence risks default, waiver, or an adverse ruling on an unopposed motion. The result is a lopsided ledger. One side generates volume for free. The other absorbs the cost at market rates. For a carrier with a duty to defend, that cost lands squarely in the loss adjustment expense column.
A meritless motion is not a free motion to refute. A hallucinated citation must be run down and rebutted before a court. A frivolous request for judicial notice must be opposed on the record. The weaker the filing, the greater the burden on the defendant to negate it.
Steep consequences for liability carriers
Several downstream effects call for attention from claims and underwriting teams alike:
- Finality is gone. Settlement is one of the most important loss-mitigation tools a carrier has. When a released claimant can manufacture Rule 60(b) arguments or simply refile the same claims in a new action, the value of “with prejudice” drops — and with it, the certainty that a resolved claim maintains its status.
- Longer claim life, unpredictable loss adjustment expenses (LAE). An increase in filings mean longer claims duration, more reserve development, and defense-cost patterns that don’t fit historical assumptions. Loss-adjustment models built for an era with natural filing governors could understate exposure on files that attract high-volume, AI-assisted activity.
- Reputational harm in the public record. In this case, the litigant placed false regulatory “sanctions” allegations into court filings and, according to the complaint, misrepresented the record to a separate arbitration forum. Defamatory content injected in a public docket under the shield of litigation privilege is difficult to remove and could bleed into parallel proceedings, as well as business relationships.
- Ancillary disputes. Motions demanding the opposing counsel’s medical records, accusations of forgery, and similar collateral filings multiply the fronts a defense team must cover — each a small, separate cost center.
None of this requires the claimant to prevail. Damage is achieved by amplifying volume and duration, not through the merits of a case.
How carriers can protect themselves now
The regulatory and judicial responses of state bills targeting AI legal tools and courts wrestling with unauthorized practice of law statutes written for humans will play out over years. The impact of this new filing dynamic is immediate. A few practical postures can help close the gap:
- Move promptly and decisively. Volume compounds. Motions to dismiss, gatekeeping requests, and vexatious-litigant remedies are more effective before a docket balloons.
- Train claims teams to spot the telltale pattern. Over-formatted filings, dense strings of statute and case citations, boilerplate rhythm, and fabricated authority are increasingly recognizable signatures of AI-generated work.
- Verify citations, document fabrications. A nonexistent case is not just a nuisance; it’s a basis for sanctions and fee-shifting. Build citation-checking into your defense workflow and document suspected AI-driven abuse contemporaneously so your records support it.
- Pursue sanctions and cost recovery when warranted.
- Tighten release and dismissal language to reinforce finality, including court-incorporated settlement terms where available.
- Revisit reserving and LAE assumptions for files exposed to pro se or AI-amplified filing volume. They might not reflect historical comparables.
Not a curiosity, a new threat
Regardless of how Nippon Life v. OpenAI is ultimately decided, the case will remain a bellwether. The tool that allows an unrepresented party to draft 60 filings for the price of a subscription is already widely used, and the cost of answering those filings falls on defendants and the carriers behind them. For excess and specialty markets — where severity is higher, risks are harder to place, and defense obligations run deep — it’s critical to treat AI-amplified litigation as an emerging driver of defense costs and claim duration that should be priced, reserved, and defended accordingly.
To learn more about the experience and expertise Liberty brings to protecting healthcare organizations, visit our healthcare liability page.
Featured insights
This website is general in nature, and is provided as a courtesy to you. Information is accurate to the best of Liberty Mutual’s knowledge, but companies and individuals should not rely on it to prevent and mitigate all risks as an explanation of coverage or benefits under an insurance policy. Consult your professional advisor regarding your particular facts and circumstance. By citing external authorities or linking to other websites, Liberty Mutual is not endorsing them.
