Every technological innovation that meaningfully changes the practice of law ultimately changes the economics of litigation. Electronic discovery transformed document production. Email reshaped the scope of discoverable evidence. Digital case management accelerated communication between lawyers, clients, and courts. Each innovation delivered increased efficiency. Yet, history has shown this rarely reduces litigation. More often, it lowers the cost of pursuing claims, expands access to the legal system, and changes the strategic balance between those bringing lawsuits and those defending them. Generative artificial intelligence may prove to be the most consequential development yet.
Much of the public conversation surrounding AI has focused on whether the technology will replace lawyers or automate traditional legal work. That debate overlooks a more immediate and consequential development already unfolding in civil litigation. Generative AI is fundamentally reducing the cost of creating legal work. Work that once required substantial attorney time can now be produced in minutes. The result is not necessarily better litigation, but considerably more of it.
As the cost of initiating litigation declines, claims that were once economically impractical to pursue may now proceed. Self-represented litigants have unprecedented access to sophisticated legal drafting tools. Plaintiffs' firms can investigate claims, develop legal theories, and prepare litigation materials with greater speed and at lower cost than ever before.
For those defending civil litigation, however, the equation is markedly different. Although AI has dramatically reduced the cost of generating legal work, it has done far less to reduce the cost of responding to it. The fundamental obligations of defense counsel have not diminished simply because the opposing party can produce legal documents more efficiently.
The widespread adoption of generative AI is unlikely to reduce the need for experienced defense lawyers. It is more likely to amplify their importance. As document production becomes increasingly commoditized, the value of defense counsel will rest less on the mechanics of drafting and more on what has always distinguished exceptional advocates such as judgment, strategy, and the ability to guide clients through increasingly complex litigation scenarios. The age of artificial intelligence may ultimately become the age in which professional judgment matters most.
The Rise of the AI-Assisted Plaintiff
In the past, practical barriers limited who could pursue civil litigation. Preparing a complaint required legal research, familiarity with procedural rules, and the ability to translate factual disputes into legally cognizable claims. Individuals who could not afford counsel frequently abandoned otherwise legitimate claims because navigating the legal system without representation proved overwhelming.
Generative AI has altered that equation. Today, a self-represented litigant can ask an AI platform to draft a complaint, identify potential causes of action, prepare discovery requests, summarize medical records, research statutes, generate deposition outlines, and draft responses to motions. Although these materials still require careful review and may contain factual or legal inaccuracies, they often resemble work product traditionally prepared by attorneys.
Recent empirical research suggests this transformation is already measurable. Professor Or Cohen-Sasson analyzed approximately 2.8 million federal civil filings and found that following the public release of generative AI tools, the percentage of federal civil cases filed by self-represented plaintiffs increased significantly. Complaints also became more citation-heavy and professionally formatted, reflecting characteristics commonly associated with AI-assisted drafting. Interestingly, the study found that these filings were not necessarily more successful. Many were dismissed earlier than traditional pro se cases, suggesting that AI improves access to the courthouse more readily than access to favorable outcomes.[i]
Researchers from the Massachusetts Institute of Technology and the University of Southern California reached similar conclusions after reviewing millions of federal court filings and docket entries. Their work likewise documents an increase in self-represented litigation following the widespread availability of generative AI and concludes that these cases place additional demands on already burdened court systems.[ii]
The implications extend beyond pro se litigants. Plaintiffs' firms are rapidly incorporating AI into virtually every stage of case development. Medical records can be summarized in minutes. Demand packages can be assembled almost immediately after representation begins. Public records, photographs, witness statements, verdict research, and legal authority can all be organized into persuasive narratives with remarkable speed.[iii] As the cost of preparing a case decline, the economics of litigation change. Claims that once may not have justified the investment of attorney time may now become economically viable. AI does not simply make lawyers more efficient. It expands the universe of cases that can realistically be pursued.
Judicial Leniency Meets AI
Federal courts and many states uniformly recognize that self-represented litigants should not be held to the same pleading standards expected of licensed attorneys. In Haines v. Kerner, 404 U.S. 519, 520-21 (1972), the United States Supreme court instructed courts to construe pro se pleadings liberally. More recently, in Erickson v. Pardus, 551 U.S. 89, 94 (2007), the Court reaffirmed that documents fueled by self-represented litigants are held to "less stringent standards than formal pleadings drafted by lawyers."
These principles serve an important purpose. They promote meaningful access to the courts by ensuring that potentially meritorious claims are not dismissed solely because an unrepresented litigant lacks legal training. Generative AI, however, complicates this longstanding framework. The traditional pro se litigant often filed pleadings or complaints that plainly reflected the author's unfamiliarity with legal procedure.
Today's AI-assisted litigant may submit professionally formatted pleadings supported by statutory citations, case law, and detailed legal argument. At first glance, these documents may resemble work prepped by experienced counsel, even though they were generated largely thorough conversational AI. Yet the procedural protections afforded to self-represented litigants remain the same. AI-generated pleadings may survive initial review more frequently, prompting additional motion practice and multiple opportunities to amend defective complaints. Even where dismissal is ultimately appropriate, the process of reaching that result may require substantially more attorney time and client expense than in years past.
The Asymmetry Problem
Perhaps the most significant consequence of generative AI is not the quality of litigation it produces, but the imbalance it creates between generating legal work and responding to it. AI has dramatically reduced the cost of creating litigation. It has not similarly reduced the cost of defending it. A plaintiff's attorney can now generate extensive interrogatories, requests for admission, document requests, deposition outlines, demand letters, and legal research in a fraction of the time previously required. AI also enable counsel to identify additional causes of action, analyze document productions, and rapidly prepare motions supported by extensive legal citations. The defense's obligations, however, remain largely unchanged.
Every interrogatory still requires legal analysis and client input. Every request for production requires a reasonable investigation, document collection, privilege review, and compliance with discovery obligations, every motion still requires counsel to verify legal authority, evaluate the governing law, and determine the most effective strategic response. The same imbalance appears during the pre-litigation claims process.
Comprehensive policy-limits demands that once arrived months into a claim now frequently appear shortly after representation begins. AI allows plaintiff's counsel to organize medical records, photographs, witness statements, timelines, verdict research, and legal authority into detailed demand packages with remarkable efficiency. Those packages often impose compressed response deadlines designed to trigger potential bad faith exposure if the insurer fails to respond appropriately.
Although AI reduces the time required to prepare these demands, it does not reduce the time necessary to investigate liability, evaluate damages, analyze coverage, interview witnesses, or advise the insurer regarding settlement strategy. The growing asymmetry may become one of generative AI's defining effects on civil litigation.
Defense Lawyers Are Essential
Generative AI is exceptionally good at recognizing patterns and generating language. Defense litigation, however, depends on judgment. Experienced defense lawyers evaluate witness credibility, assess litigation risk, recognize factual inconsistencies, advise insureds through difficult decisions, negotiate resolutions, and develop strategies tailored to each case. They determine whether a particular deposition advances the client's interests, whether summary judgment is likely to succeed, whether settlement creates future exposure, and how jurors respond to competing narratives. Those decisions cannot be delegated to an algorithm.
Nor can AI replicate the trust developed between counsel, insureds, claims professionals, and the courts, Successful defense practice depends as much upon credibility, professional judgment, and experience as it does upon legal research or drafting ability.
None of those diminishes AI's value. To the contrary, defense firms should embrace generative AI to summarize deposition transcripts, organize medical records, identify inconsistencies in testimony, prepare chronologies, suggest legal authorities, and produce first drafts of routine documents. Used responsibly and within secure environments that protect privileged information, these tools can significantly improve efficiency. The greatest benefit of AI is that it allows lawyers to spend less time producing documents and more time exercising judgment.
Looking Ahead
Generative AI is unlikely to reduce litigation. If current trends continue, it will almost certainly increase it. Lower barriers to entry will encourage more self-represented litigants. Plaintiff firms will continue leveraging AI to investigate claims, prepare demand packages, and generate pleadings more efficiently. Discovery will become broader. Motion practice will become more frequent. The volume of litigation confronting defense counsel is likely to grow. Ironically, these developments do not diminish the importance of defense lawyers. They reinforce it.
As document generation becomes increasingly commoditized, the legal professional's competitive advantage will no longer live in drafting the fastest brief or producing the longest discovery responses. It will live in providing what AI cannot, such as sound judgment, strategic thinking, ethical decision-making, credibility, and the ability to counsel clients through uncertainty.
In the age of generative AI, the most valuable defense lawyer will be the one who known which documents matter, what arguments advance the client's objectives, and when experience, not automation, should guide the next decision.
___________________________
[i] Or Cohen-Sasson, The AI-Powered Pro Se: Assessing AI's Impact on Self-Represented Litigation (2026), available at https://arxiv.org/abs/2605.29493
[ii] Yoon-Ho Alex Lee et al., Generative AI and Federal Civil Litigation (MIT & USC Working Paper 202) available at SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6766859
[iii] Chupryna, Roman & Serhii Leleko. How Artificial Intelligence Transforms Legal Services in 2026.SPD Technology. May 24, 2024. https://spd.tech/artificial-intelligence/how-artificial-intelligence-transforms-legal-services/
