The Autonomous Brief: Navigating the Ethical Line Between AI Assistance and the Unauthorized Practice of Law (UPL)
Related Posts
-
The Hybrid Future of International Commercial Disputes: Where Human Expertise Meets AI Efficiency with NexLaw.ai
-
The Global Courtroom: How International ODR and AI Are Revolutionizing Cross-Border Disputes
-
Minimizing Court Backlogs: The Preventative Power of AI in Early Dispute Resolution and Mediation
The Autonomous Brief: Navigating the Ethical Line Between AI Assistance and the Unauthorized Practice of Law (UPL)
The conversation around Legal AI is rapidly evolving. The familiar generative AI that responds to a lawyer’s prompt is being joined by a powerful new technology: agentic AI. Unlike its predecessor, agentic AI is goal-driven. It can independently plan and execute a series of tasks to achieve a stated objective. An agentic AI might be instructed, “Draft a complete motion to dismiss based on the attached complaint, conduct the necessary research and file it with the court.” This leap in autonomy presents incredible opportunities for efficiency, but it also brings a critical ethical question into sharp focus for U.S. lawyers: at what point does an AI’s autonomous work cross the line from sophisticated assistance into the Unauthorized Practice of Law (UPL)?
As legal professionals navigate this new frontier, understanding the ethical guardrails is paramount. This guide explores the emerging ethical challenges of agentic AI and explains how a professionally designed platform like NexLaw AI helps lawyers harness this new power responsibly.
The Problem? The "Black Box" and the Non-Delegable Duty of Judgment
The core of UPL rules is the principle that legal judgment, the application of legal principles to a specific set of facts to provide advice, can only be exercised by a licensed human lawyer. The risk with a highly autonomous or “agentic,” AI is that it can become a “black box,” performing complex legal tasks without the lawyer’s direct supervision or understanding. This implicates several core duties:
- The Duty of Competence (Rule 1.1):Â A lawyer cannot be competent in their representation if they do not understand the reasoning behind the legal arguments they are submitting.
- The Duty of Supervision (Rules 5.1 & 5.3): A lawyer is responsible for all work done on their behalf, whether by a paralegal or an AI. They cannot simply delegate a task and accept the output without review.
- The Duty to Avoid UPL (Rule 5.5):Â A lawyer must not assist a non-lawyer (which includes an AI) in the unauthorized practice of law.
The State Bar of California’s guidance on generative AI already emphasizes these risks, warning that relying on AI outputs without verification can violate the duty of competence. As AI becomes more agentic, these warnings become even more critical.
The AI Advantage? The "Glass Box" Model of Augmented Intelligence
The key to ethically using agentic AI is to ensure it operates not as an autonomous decision-maker, but as a highly advanced, transparent assistant under the lawyer’s complete control. This is the “glass box” model, where the AI’s processes are visible and its outputs are subject to human review and approval at every key stage. A professionally designed AI Legal Assistant like NexLaw AI is built on this principle.
You can give the AI a high-level goal, like “Prepare for the deposition of Witness X,” but the platform breaks this down into a series of discrete, reviewable tasks (e.g., “Summarize past testimony,” “Identify inconsistencies,” “Draft potential questions”).
The AI does not proceed from one task to the next without human approval. It presents its findings at each stage, requiring the lawyer’s input and judgment before moving forward.
Every piece of information the AI uses and every step it takes is tracked and linked to a source, creating a clear audit trail that allows the lawyer to understand and defend the final work product.
See NexLaw in Action
Start your free trial and kick off your legal AI journey with a personalized demo
*By submitting the form, you agree to the Terms of Service and Privacy Policy
A Step-by-Step Workflow for Ethical Agentic AI Use
Let’s imagine a lawyer using NexLaw AI’s future agentic capabilities to prepare a Motion for Summary Judgment.
Step 1: The Lawyer Sets the Goal
The lawyer initiates the process: “Begin preparation of a Motion for Summary Judgment in the Acme v. Globex case on the issue of tortious interference.”
Step 2: The AI Proposes a Plan (Task 1)
The AI does not start writing. It presents the lawyer with a proposed plan:
- Analyze the complaint and answer to define the elements of tortious interference under Illinois law.
- Scan the discovery record in ChronoVault to identify all evidence related to each element.
- Conduct research for controlling case law in the Seventh Circuit on this issue.
- Propose a structure for the motion. The lawyer reviews and approves this plan.
Step 3: The AI Executes and Presents Findings (Tasks 2 & 3)
The AI executes the first few steps and then presents a report to the lawyer: “Here are the key pieces of evidence supporting our motion and here are the three most relevant appellate cases. Please review and confirm.” The lawyer reviews the evidence and the cases, using their judgment to select the strongest points.
Step 4: The AI Drafts, The Lawyer Refines (Task 4)
Based on the lawyer’s approved selections, the AI generates a first draft of the motion. The lawyer then takes over, refining the persuasive language, strengthening the arguments and ensuring the brief aligns with their unique voice and style. The final work product is a synthesis of AI’s efficiency and the lawyer’s irreplaceable judgment.
Case Study: The "Autonomous" but Supervised Conclusion: The Lawyer as Commander, The AI as Agent
Legal AI nor Agentic AI is not a threat to the legal profession; it is the logical evolution of the legal assistant. It represents a move from a tool you have to operate manually to an agent you can command. However, the lawyer must always remain the commander. By using a platform that is built on the principles of transparency, verifiability and human-in-the-loop supervision, U.S. lawyers can ethically harness the power of autonomous AI to practice at a higher, more strategic level than ever before.
Ready to explore the future of legal automation responsibly?
Book a Personalized Demo to discuss how NexLaw AI is building the future of ethical and effective agentic AI.
Explore Our Plans and partner with a platform that respects the irreplaceable role of the lawyer.
Interested In Features Like This?
Receive complimentary access to our resources and a personalized live demo tailored to your needs.
