The New Pro Se: Generative AI and the Surge in Federal Civil Self-Representation

A comprehensive study of 2.8 million federal civil filings reveals that generative AI has driven pro se (self-represented) litigation rates from 11.33% to 16.94% since public AI access became widespread. While AI-flagged complaints show higher citation density and attract first-time filers, they paradoxically suffer worse outcomes with higher dismissal rates, raising critical questions about whether AI-assisted legal drafting improves access to justice or merely creates the appearance of formality.

The proliferation of generative AI tools has fundamentally altered the landscape of federal civil litigation, creating an unexpected consequence: a measurable surge in self-represented plaintiffs. This 5.61 percentage-point increase in pro se filing rates, validated across multiple robustness checks, suggests that AI has lowered perceived barriers to legal participation for individuals without formal legal training. The effect concentrates most heavily in Civil Rights and Statutory cases, indicating that AI tools are particularly attractive to plaintiffs navigating complex legal domains where professional representation remains expensive or inaccessible.

The stylometric analysis revealing 13.9% of post-GenAI complaints as AI-flagged provides tangible evidence of this shift. These AI-assisted complaints demonstrate higher citation density and disproportionately originate from first-time filers and across geographically dispersed regions, suggesting that AI democratizes drafting across demographic boundaries. Notably, the data hints at modest increases among female plaintiffs, indicating AI may reduce traditional barriers to legal participation for underrepresented groups.

However, the study's central finding undercuts optimistic narratives about AI-enabled access to justice: AI-flagged complaints face higher dismissal rates and terminate earlier in procedural phases, with no evidence of improved win rates. This disconnect between legal formality and legal efficacy reveals a critical limitation of current generative AI systems. They excel at producing formally correct documents that satisfy superficial requirements but fail to construct legally persuasive arguments tailored to judicial expectations and case-specific nuances. This creates a troubling scenario where AI amplifies filing volumes while degrading case outcomes, potentially straining already-burdened courts with frivolous filings and burdening self-represented parties with false hope.

• →Pro se filing rates in federal civil litigation increased 5.61 percentage points after generative AI became publicly accessible, rising from 11.33% to 16.94%.
• →Stylometric analysis identifies 13.9% of post-GenAI complaints as AI-flagged, predominantly drafted by first-time filers rather than repeat litigants.
• →AI-assisted complaints are more citation-dense but suffer higher dismissal rates and worse procedural outcomes than human-drafted filings.
• →The surge in AI-drafted complaints suggests tools democratize legal participation across demographics, including a modest increase among female plaintiffs.
• →AI enhances formal legal drafting appearance without improving substantive legal efficacy, creating a gap between document quality and litigation success.