Our Take
Access to legal filing tools is not the same as access to justice, and courts are discovering the cost of that gap falls on judges, not vendors.
Why it matters
The flood of AI-drafted filings is forcing federal courts to absorb triage work and legal review that was previously either filtered by attorney gatekeeping or handled by pro bono counsel. Lawmakers now face a real liability question: when a chatbot writes a meritless complaint, who is liable for the harm to the defendant and court system?
Do this week
If you ship legal-advice tools or LLM-based document drafting for courts or self-represented litigants, audit your terms of service and insurance coverage now for liability exposure when outputs fail substantive review.
Filings without lawyers surge, but case results stay flat
Federal magistrate judge Maritza Braswell in Colorado reports that the volume of pro se (self-represented) filings in her chambers has more than doubled since 2023. Braswell attributes the jump directly to AI tools, which now enable individuals to draft complaints, motions, and other legal documents without hiring counsel.
The numbers are striking: more than twice the volume in three years. Yet the data shows no improvement in win rates or case outcomes for these filers. The increase in volume appears to reflect access to a filing tool, not access to competent legal representation.
Judges are now confronting a secondary problem: the volume of AI-drafted filings requires judicial resources to sort signal from noise. Courts are wading through documents that may meet procedural minimums but lack substantive merit, placing the burden of review on the bench rather than on counsel or filers themselves.
The liability question now falls on lawmakers
Two distinct problems are colliding. First, defendants face exposure to meritless claims that consume legal and financial resources to dismiss, even when outcome is clear. Second, courts face caseload inflation driven not by genuine disputes but by friction-free filing.
Judges are beginning to ask hard questions about the rights and duties of chatbots when they function as ersatz counsel. But the sharper legal question is backward-facing: who bears the cost when an LLM produces defective legal advice? Vendors have largely avoided liability by disclaiming legal advice. But as these tools migrate from exploration to routine use, lawmakers will need to resolve whether vendors should carry insurance, duty of care, or explicit liability caps for outputs that damage litigants.
The third-order problem is systemic: if courts must absorb the cost of reviewing AI-drafted filings, the judiciary becomes an unpaid quality-control layer for products sold or distributed for free. That is a silent subsidy of vendor business models.
Audit your legal-drafting disclosure and liability insurance
If your product helps users draft legal documents or correspondence intended for court filing, review your terms of service and disclaimers now. Confirm that you explicitly disclaim any warranty that outputs are legally sufficient or accurate. Check whether your errors and omissions insurance (or your customer's insurance) covers output liability when your tool produces a complaint that leads to sanctions, dismissal with prejudice, or a default judgment against the filer.
Do not assume that a simple "not a lawyer" disclaimer is sufficient protection. Courts are beginning to push back on the idea that vendors can shed responsibility simply by labeling the tool as advisory. If the tool is marketed as capable of generating court filings, disclosure should match capability, not minimize it.