MyPillow Motion Filings with AI-Fabricated Citations and Repeat Offense
Attorneys filed motions with AI-fabricated citations to nonexistent cases, were sanctioned, then filed again with another flawed citation despite the prior sanction.
What happened
In 2025, attorneys for Mike Lindell filed a Denver motion containing approximately 30 defective citations produced by an AI system, including citations to nonexistent cases. Judge Nina Wang fined the attorneys $3,000 each in July 2025. Despite this sanction, in May 2026, attorney Christopher Kachouroff and his firm were sanctioned again—$5,000—after filing another motion with a flawed AI-generated citation in a later filing. The repeat offense occurred without any intervening approval gate that would have required human verification of the citation accuracy before filing.
What the agent did
The attorney submitted motions containing AI-generated fabricated citations to the court docket without mandatory human approval of the citation verification results.
The irreversible effect
Fabricated citations entered into the court record; attorney sanctions ($3,000 in first instance, $5,000 in second) imposed; damage to professional reputation; potential bar discipline exposure.
Root cause
No governance control requiring approval of the irreversible filing action. Filing with the court lacked a mandatory approval gate that would trigger on every submission and require human review of citation verification before docket submission. The first sanction was not architecturally built into the system to prevent recurrence through automated governance.
How a maker-checker control would have refused it
high_risk_requires_gate: Filing to court is published as a high-risk skill; the proxy categorically refuses direct submission outside a governed flow with a preceding approval gate. The gate fires on every submission (including resubmissions), forcing human review of the citation verification record before docket entry is permitted.
Runnable reproduction
This incident ships as a runnable scenario in the open-source repository. Point the enforcement engine at the policy and watch the action get refused, with the refusal written to a signed audit record.
examples/mypillow-ai-brief-fake-citations-repeat
Accuracy and corrections
This entry describes a publicly reported incident and is compiled from the primary sources listed above. Where an account is a legal allegation rather than an established finding, the entry labels it as such. Summaries can still contain errors. If you can document a correction, email hello@makerchecker.ai and we will review and correct it, with the change noted, within 14 days.
See it for yourself
Reading is one thing. Watch it block an agent.
One command starts the demo: an agent stopped from signing off its own work, and the signed evidence file an inspector can check for themselves.
Designed against the rules your auditors already enforce.