Artificial intelligence may be making the practice of law faster. However, a series of recent New York decisions is making something else crystal clear: AI does not replace a lawyer’s obligation to actually know the law, verify the law, and take responsibility for what is filed in court.
The September 2026 decision in Youssian v. Village of Kings Point, 2026 NY Slip Op 26150, may be one of New York’s strongest warnings yet.
The case did not involve a lawyer simply misspelling a citation or getting a page number wrong. The court found that the papers contained fabricated or altered quotations attributed to real cases and relied upon supposed authorities in support of substantive legal arguments. At LawByDavid.com, we stay current on changes in the law and independently verify the cases and legal authorities we rely on, so our clients receive thorough, informed, and effective representation.
The consequences were severe: the court struck both the original petition and amended petition, dismissed the Article 78 proceeding, sanctioned counsel $5,000, and ordered counsel to pay a combined $10,000 in attorneys’ fees and costs to the respondents.
That should get the attention of lawyers and clients.
What Happened in Youssian v. Village of Kings Point?
The petitioners filed a substantial case: a 64-page petition together with a 46-page memorandum of law.
Then opposing counsel checked the authorities. They alleged that the papers contained numerous false legal citations and quotations generated through artificial intelligence. Instead of immediately conceding the problem, petitioners’ counsel, Thomas A. Bizzaro, Jr., Esq., initially sent an eight-page response disputing the accusations and making threats.
The court later described his initial reaction as essentially “doubling down.” Only afterward did he acknowledge that the filing contained what he called “AI-citation errors.”
At a hearing, Bizzaro explained that his firm used Lexis and Westlaw and that an AI program called Spellbook AI was embedded into Microsoft Word. According to the decision, he said he had been told Spellbook improved writing and detected drafting problems and that it was not supposed to provide legal authorities or citations.
According to Bizzaro, he reviewed the petition. But after accepting Spellbook’s changes, he did not perform another citation check.
There is a plethora of AI companies being used for by consumers for work and for personal tasks such as OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini, Microsoft’s Copilot, Perplexity, Harvey, Lexis+ AI (protege), and Westlaw Precision AI.
Whether the software is Spellbook, ChatGPT, Claude, Gemini, Microsoft Copilot, Perplexity, Harvey, Lexis+ AI, Westlaw Precision AI, or the next legal AI platform released months from now, lawyers cannot outsource professional judgment to software.
Artificial intelligence can assist with legal work, but it is nowhere near a substitute for a lawyer. When the stakes involve something as serious as your legal rights or your health, relying on AI alone can be dangerous. AI may be useful for general information, but important legal or medical decisions should be made only after consulting a qualified professional.
The court ordered Bizzaro to pay:
$5,000 to New York’s Lawyers’ Fund for Client Protection; $8,000 toward attorneys’ fees for 22 Liam Properties LLC; and $2,000 toward attorneys’ fees for the Village respondents.
The Most Dangerous AI Mistake Is Not Always a Fake Case
People often think an “AI hallucination” means it invents a completely fictional lawsuit. That certainly happens. But Youssian illustrates something potentially harder to detect. The AI can give you a real case and then attribute language to that case that does not exist.
The petition cited legitimate New York authorities, including Pell v. Board of Education and Douglaston Civic Association v. Klein. But the Nassau County Supreme Court found that quotations or propositions attributed to several authorities were fabricated, altered, or unsupported by what those decisions actually held.
That can be more dangerous than a completely fictional case.
someone still has to read the case.
At David Ilganayev Law, PLLC, we stay current on changes in the law and independently verify the cases and legal authorities we rely on so our clients receive thorough, informed, and effective representation.
New York Courts Are Building an Actual Body of AI Sanctions Law
Youssian is be noteworthy on its own, but it is much more significant because it joins a rapidly growing series of New York decisions addressing fabricated AI legal research.
In Landberg v. City of New York, 2026 NY Slip Op 03935, the Second Department encountered an appellate brief containing citations to nonexistent cases, fictitious quotations supposedly from New York’s highest court, and inaccurate descriptions of what real decisions held.
The court imposed an $8,000 sanction on the attorney and another $2,500 sanction on the law firm.
In Deutsche Bank Natl. Trust Co. v LeTennier, 2026 NY Slip Op 00040 (3d Dept Jan. 8, 2026), the Third Department found that the defendant had submitted at least 23 fabricated legal authorities across five filings during the pendency of the appeal, along with misrepresentations of the holdings of several real cases as dispositive in his favor.
The Court imposed $7,500 in sanctions against defendant’s counsel; $5,000 for the fabricated authorities and $2,500 for prosecuting a frivolous appeal; and an additional $2,500 against the defendant, for a combined $10,000 payable within 60 days under 22 NYCRR 130-1.1.
AI Legal Advice Can Be Even More Dangerous for Clients
The same warning applies to people who increasingly use Claude, ChatGPT or other AI systems as substitutes for lawyers. AI can be excellent for explaining terminology, helping someone organize facts, or giving a general overview of an unfamiliar subject. The danger begins when someone starts relying on an AI answer to make actual legal decisions.
A lawsuit may turn on a filing deadline, the precise wording of a contract, whether service was proper, whether an argument was preserved, which appellate department controls, whether a statute changed, or whether a case cited by the AI actually says what the AI claims it says. An artificial-intelligence system can give an extremely polished and confident answer while getting one of those critical details wrong.
That mistake can become expensive very quickly. A defective filing may need to be redone. An argument can be waived. A deadline can expire. A bad admission can become part of the record. A defective notice can delay a case for months. In some situations, the mistake cannot be repaired at all.
That is why the cost of obtaining competent legal advice should be compared not only with the cost of an AI subscription or inexpensive document preparation, but with the potential cost of getting the law wrong.
But there is a significant difference between using technology to assist legal work and allowing technology to become the legal work.
At David Ilganayev Law, PLLC, technology is a tool used to support legal work, not a substitute for doing it. We review the record, stay current on changes in the law, verify the authorities we rely upon and remain responsible for the arguments made on behalf of our clients.
When your business, home, money, freedom or lawsuit is at stake, that distinction matters.
📞 Call LawByDavid.com at (718)-200-2866 or Contact Us now for a confidential case review and strong legal defense.