Matthew Brown (Matthew Brown Companies) has posted Agentic Artificial Intelligence and Pro Se Federal Litigation: A Participant Observer Case Study of SEC v Brown on SSRN. Here is the abstract:
This Article presents a participant-observer case study of artificial intelligence used across the lifecycle of an active federal securities-enforcement action. In SEC v. Matthew Brown and Matthew Brown Companies, LLC, the individual defendant proceeded largely pro se while using a network of commercial language models, retrieval systems, application programming interfaces, document repositories, and specialized agents to support research, drafting, discovery, deposition preparation, record analysis, adversarial testing, and remedies briefing. The district court granted the Securities and Exchange Commission partial summary judgment on liability; remedies remained pending as of September 26, 2026. The case therefore does not establish that artificial intelligence can replace counsel, and the adverse liability result supplies an essential counterweight to any account of productivity. It does, however, provide unusual evidence about how AI can reduce analytical latency, externalize litigation memory, permit asynchronous production, and unbundle some forms of legal labor from the professional judgment with which they have traditionally been sold. The Article distinguishes docket facts from the author’s recollection, describes the architecture and controls used, identifies consequential failures, and develops five provisional findings. First, the relevant unit of analysis is increasingly a human-machine workflow rather than a single chatbot. Second, orchestration and verification may matter more than raw model performance. Third, AI can expand the feasible boundary of self-representation without equalizing institutional power. Fourth, low software prices obscure substantial costs in operator time, technical skill, and error risk. Fifth, access-to-justice gains and professional responsibility risks arise from the same capability: inexpensive production of plausible legal work at scale. The Article closes with a governance framework for AI-assisted litigation and a research agenda for more systematic study.
Recommended!
To receive new posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.
Lawrence Solum