OSINT Case Study: AI-Generated Filings in Wright v. Ramsey County
Date: November 24, 2025
This page documents how an incarcerated pro se litigant, Antonio Dupree Wright, is using AI-generated pleadings to advance a federal civil-rights case against Ramsey County and other defendants in the District of Minnesota. The focus here is narrow: not on the substance of his underlying criminal charges, but on the observable fingerprints of large-language-model text in his filings and how the federal court has responded to those filings procedurally.
Wright’s Amended Complaint in Wright v. Ramsey County, et al., Case No. 0:25-cv-02502-JRT-DLM, bears all of the expected hallmarks of AI-assisted drafting. The document chains long, formal sentences together with commas and semicolons; groups statutes into dense “citation clusters” that read like a language-model synthesis; repeats stock phrases such as “amongst others” and “Plaintiff incorporates by reference all preceding and succeeding paragraphs”; and uses near-identical paragraph structures across unrelated sections. The supplemental filing later in the docket amplifies these traits even further.
Despite these stylistic markers, the federal court did not reject his work as incoherent or frivolous. Instead, Magistrate Judge Douglas Micko accepted the Amended Complaint as the operative pleading and entered a briefing order mooting two pending motions to dismiss. Defendants already served were directed to answer the new complaint by a fixed deadline. In other words, a document that appears to have been drafted with the heavy assistance of an AI system has successfully carried a pro se prisoner past the earliest choke point where most civil-rights plaintiffs—represented or not—lose their cases.
From an OSINT perspective, this is significant for several reasons. First, it confirms that federal courts are willing to treat AI-generated pleadings as valid if they are factually grounded and procedurally proper, even when the prose is obviously machine-shaped. Second, it provides a live example of an incarcerated person using AI as a compensatory tool against the structural advantages of institutional defendants and county counsel. Third, it demonstrates that the presence of AI in a filing does not, by itself, doom a civil-rights action to dismissal—what matters is whether the complaint gives the court enough factual content to work with.
Wright’s strategy parallels my own use of AI in Strickland v. Ramsey County, et al.: using language models to organize complex timelines, surface legal theories, and keep pace with a high volume of defense filings, while retaining human control over facts, exhibits, and final edits. The difference is that his work appears far less edited, which makes the AI fingerprints easier to see. That visibility turns his case into a useful case study for how courts are actually treating AI-assisted litigation in practice, beyond the panic headlines and ethics memos.
Summary in Context
Wright’s progress through the early stages of federal litigation—despite, and in some ways because of, his reliance on AI—undercuts the narrative that AI-generated legal text is inherently incompatible with serious civil-rights work. When used as a tool rather than a substitute for evidence, AI can help a disabled or resource-constrained litigant keep up with large institutional defendants. His case also intersects with a separate concern documented on this site: the presence of conflicting Registers of Actions in his underlying Ramsey County case, showing the same broken index-number pattern that appears in my dockets and in Scheffler. Together, those two threads—AI-assisted pleading and unstable court records—show how pro se litigants are adapting to, and exposing, the structural weaknesses of the systems arrayed against them.
This page will be updated as Wright’s case develops, particularly if future orders reference the quality or origin of his pleadings, or if the court adopts any explicit standards for AI use in filings. For now, his docket stands as a concrete, real-world example of AI being used as a survival tool in impact litigation rather than as a gimmick or shortcut.