Incident Overview
On July 30, 2025, Boies Schiller Flexner, a prominent litigation firm, filed an appellate brief in Bixler v. Church of Scientology International. The brief was reportedly prepared with assistance from generative AI tools. Shortly after submission, opposing counsel identified significant errors in the citations, including a nonexistent case and mischaracterized legal authorities. The firm acknowledged the mistakes and took responsibility for failing to verify the citations, which were attributed to partner John Kucera.
The errors included a fabricated case citation and inaccuracies in the presentation of existing legal precedents. These mistakes raised concerns about the reliability of AI-assisted legal drafting workflows, particularly in high-stakes litigation. The court denied the firm’s request to withdraw the original brief and later considered imposing monetary sanctions as a consequence of the citation failures.
The incident occurred within the context of Bixler v. Church of Scientology International, a high-profile lawsuit involving allegations against the Church of Scientology. The official appellate docket for the case provides further documentation of the citation disputes and the firm’s response.
This case stands out as one of several 2025 incidents involving prominent law firms where AI-assisted drafting tools introduced material errors into legal filings. The harm in this instance was primarily reputational, affecting the firm’s credibility, the integrity of the judicial process, and the trust of clients relying on AI-enhanced legal work.
The AI systems implicated in this incident include Boies Schiller Flexner’s AI-assisted legal drafting workflow and broader generative AI tools used for legal research and citation verification.
Possible Root Cause
The root cause of the citation errors in the Boies Schiller brief appears to stem from overreliance on AI-assisted drafting tools without adequate human oversight. Generative AI models, while capable of synthesizing legal arguments and citations, are prone to hallucinations—generating plausible but factually incorrect information. In this case, the AI system produced a nonexistent case citation and misrepresented established legal authorities.
Partner John Kucera accepted responsibility for the failures, indicating that the firm’s internal review process did not catch the errors before submission. This suggests a breakdown in the workflow where AI-generated content was not thoroughly vetted by legal experts familiar with the nuances of the case. The incident highlights the need for robust verification protocols when integrating AI into legal drafting, particularly in litigation involving complex or high-profile matters.
The use of AI in legal research and drafting is increasingly common, but the Bixler case demonstrates the risks when such tools are deployed without safeguards. The firm’s acknowledgment of the errors underscores the importance of human judgment in validating AI-generated outputs, especially in contexts where inaccuracies can have significant legal and reputational consequences.
The Church of Scientology, as the opposing party, likely capitalized on the errors to challenge the firm’s credibility, further complicating the litigation process.
Takeaways
Church of Scientology incident serves as a cautionary example of the potential pitfalls of AI-assisted legal drafting when human oversight is insufficient. Key takeaways from this case include the necessity of implementing rigorous review processes for AI-generated legal content, particularly in high-stakes litigation where accuracy is non-negotiable.
Human lawyers must verify all citations, legal reasoning, and factual claims before submission to the court. This includes cross-referencing case law, ensuring proper formatting, and confirming the existence of cited authorities.
Second, the case highlights the reputational risks associated with AI errors in legal filings. Even if the harm is primarily epistemic—such as misinformation or fabricated claims—the fallout can damage a firm’s standing with clients, courts, and the public. The Church of Scientology’s ability to exploit the citation errors for strategic advantage demonstrates how such mistakes can be weaponized in adversarial proceedings.
Third, the incident underscores the importance of transparency in AI-assisted legal workflows. This transparency can mitigate skepticism from opposing counsel and judges, who may otherwise question the integrity of the submission.
Finally, the Bixler case contributes to a growing body of evidence about the limitations of current generative AI systems in legal contexts. While these tools can enhance productivity, they are not infallible and require careful integration into existing quality control measures. Law firms must balance efficiency with accuracy, ensuring that AI serves as an assistant rather than a replacement for human expertise.
As more firms incorporate generative AI into their workflows, the Bixler case provides a real-world example of the consequences when such tools are misused or inadequately supervised.
Sources
Incident summary from the AI Incident Database, used under CC BY-SA 4.0.