In a quiet courtroom drama that could pass for a cautionary tale about technology and justice, Fa’alagilagi Meni-Siliga is not a statistic or a legal footnote. She is a U.S. Postal Service carrier who watched her family’s home in Carson, Los Angeles County, slip away after a fire left it uninhabitable in 2020. The house was supposed to be repaired, but years later Meni-Siliga and her husband had drained their savings and retirement accounts, declared bankruptcy, and lost the property to foreclosure. In July 2024, she filed a lawsuit naming her contractor, the contractor’s insurance company, a public adjuster, and her own homeowners insurer, State Farm, as defendants. Her lawyer, Eric Khodadian, later made a stunning discovery: some of the legal filings State Farm’s lawyers submitted on its behalf were riddled with artificial intelligence hallucinations—nonexistent case law, fabricated quotes, and invented holdings. It was an embarrassing lapse for one of the nation’s largest insurers, but more importantly, it threatened to undermine a case already freighted with heartbreak and financial ruin. The incident has become a vivid example of what happens when the legal profession’s growing reliance on generative AI collides with the lives of ordinary people who simply wanted their home rebuilt.
The human story behind the legal scandal begins long before any lawyer opened an AI tool. In 2020, a fire tore through Meni-Siliga’s home in Carson, making it uninhabitable and forcing her family into an agonizing limbo. While a contractor worked on the property, a storm caused additional water damage, compounding the original disaster. What followed was a frustrating cycle of delays, particularly around getting State Farm’s approval for repairs. According to court filings, the work was never completed, and the family’s financial foundation crumbled beneath them. Meni-Siliga described the devastation in her own words: “My family and I have lost forever our family home. We do not come from money… It took many years and sacrifice for our family to purchase our home.” That sacrifice was not abstract. It meant years of careful saving, of choosing stability over indulgence, of building a place where her family could feel safe and rooted. By the time she filed her lawsuit, the dream had already been foreclosed upon, quite literally. The case is set to go to trial in October, but no courtroom victory can restore the years of lost memories, the security of a family homestead, or the simple dignity of having a place to call home.
When State Farm initially responded to Meni-Siliga’s complaint, it did not admit any failure. Instead, the insurer argued that negligence by the plaintiff herself, her contractor, and her public adjuster had caused the injuries and damages she suffered. State Farm also asserted that she had failed to bring her legal action within one year of the original loss or damage, a procedural defense that would have been cold comfort to a family already devastated by the slow unraveling of their lives. It was during this litigation that Khodadian began scrutinizing State Farm’s filings and found what he described as “cases that do not exist, quotes that do not exist, and holdings that do not exist.” He shared his findings at an Aug. 7 conference in Los Angeles Superior Court. According to Khodadian, a State Farm lawyer, Kenneth Katel of the firm Musick, Peeler & Garrett, was “visibly enraged” afterward and followed him and his co-counsel down the hall, “repeatedly demanding to see the list” of mistakes. Katel later apologized to the court and to the plaintiff, explaining that his co-counsel had told him she used AI to help prepare some filings. “I was not aware that AI had been used, but as lead trial counsel I accept full responsibility for our filings,” Katel wrote. He disputed Khodadian’s account of the hallway confrontation, though he conceded, “it is accurate to say I was upset given the seriousness of the allegations.” The lawyer who actually used the AI tool, Jacquelene Robinson, confirmed that seven case citations across eight filings “simply didn’t exist,” along with incorrect case titles and quotes that could not be found in the cited authorities. She said she had used an AI tool called Irys, believing it was connected to her law firm’s subscription to Westlaw and would “perform an internal cite check.” She was mistaken.
This is not an isolated anomaly but a symptom of a rapidly changing profession. According to a recent survey, 41% of law firms and 47% of corporate legal departments are now using generative AI in some capacity. With that adoption has come a steep rise in what are euphemistically called legal hallucinations—confidently rendered but entirely fabricated citations and quotations. Daniel Ho, a law professor at Stanford University who has researched the issue, co-authored a paper last year finding that some companies “overstated” their legal research tools’ ability to guarantee hallucination-free citations. A Paris-based legal researcher’s database of legal hallucinations around the world has cataloged 1,922 cases so far, with most occurring in the United States. Perhaps the most infamous example came last year, when a California attorney was fined a record $10,000 for using ChatGPT to draft an opening brief that contained 21 made-up quotes out of 23. These incidents reveal a fundamental tension: AI tools can draft documents at astonishing speed, but they cannot yet consistently distinguish between what is real and what merely sounds plausible. In the high-stakes world of litigation, where a single fake citation can waste hours of court time and potentially derail a case, the consequences are far from academic. For Meni-Siliga, the AI-generated falsehoods in State Farm’s filings were not just a procedural problem; they were another delay, another indignity, another sign that the system meant to help her could not be trusted to tell the truth.
The emotional weight of this case extends beyond the legal malpractice. Meni-Siliga’s family has already paid the highest price a family can pay in this situation: the loss of the home they worked their entire lives to establish. She is a working woman, not a corporation with bottomless legal resources, and the fire that started this nightmare was not her fault. Yet she has endured years of bureaucratic inertia, financial collapse, and now the surreal experience of watching lawyers submit phantom legal authorities in opposition to her claims. State Farm, for its part, has said it is reviewing what happened. Spokesperson Tom Hartmann stated: “State Farm expects its external counsel to conform to the highest level of ethical standards and professionalism, including confirming the accuracy of all legal filings.” But Hartmann did not answer questions about how many external law firms represent State Farm or whether the company asks about their AI policies before hiring them. Katel noted that his firm has updated its policy on AI use, though he provided no specifics. Those statements may satisfy institutional requirements, but they do little to reassure a woman who has already lost everything. An apology, however sincere, cannot rebuild a foreclosed home or restore drained retirement savings. It can only acknowledge that the system failed her again, this time through a technology no one fully understands but everyone seems eager to embrace.
The broader lesson is not that lawyers should abandon AI, but that they must treat it with the same caution they would apply to any junior assistant: verify everything, cite only what you have personally read, and take responsibility when mistakes are made. Generative AI is an extraordinary tool for brainstorming, summarizing, and organizing information, but it is not a source of legal truth. The case against State Farm is a reminder that hallucinations are not harmless glitches; they are potential acts of injustice. Courts may need to adopt clearer rules requiring lawyers to certify that their citations have been checked against primary sources, and law firms must create cultures where human judgment always overrides algorithmic confidence. For Meni-Siliga, the October trial will be a moment of reckoning. She will walk into court as a plaintiff who has lost her home, her savings, and her sense of security, seeking accountability from corporations and contractors who she believes let her down. The AI scandal swirling around State Farm’s filings has already taken a toll on the credibility of its defense, and it should serve as a wake-up call to the entire legal profession. Technology can make law faster and cheaper, but it cannot make it more compassionate. Justice still depends on human attention, human empathy, and the willingness to admit that no machine, no matter how sophisticated, can fully understand what a family loses when it loses the roof over its head. As the legal world embraces artificial intelligence, it must not forget that behind every case file is a person whose life is not hypothetical. For Meni-Siliga, her family’s story is not a cautionary tale about algorithms—it is the simple, tragic truth that home is irreplaceable, and no tool or apology can change that.