As artificial intelligence moves deeper into courtrooms, law offices, and intake centers across the United States, one attorney-turned-technologist is making the case that accountability in law cannot be outsourced to an algorithm.
Sophia Snomi, founder of VerdictAI, has launched a scholarly book titled “The Supervised Attorney: Professional Liability, Legal AI, and the Boundaries of Machine-Assisted Practice,” published by Lambert Academic Publishers. The book is now available on Amazon and other major e-commerce outlets.
Snomi brings direct practice experience to the subject. Before founding VerdictAI, she served as Legal Counsel at Chocolate City Group, one of Africa’s largest entertainment companies, where she handled commercial agreements and international licensing matters.
She later worked on regulatory compliance and governance at Nigeria’s Border Communities Development Agency before earning admission to the Texas Bar and joining Thomas J. Henry Law, one of the state’s largest personal injury firms. Working inside a high-volume litigation practice gave her firsthand exposure to the operational and ethical gaps that AI systems are now being deployed to fill.
That experience is what sets the book apart from existing legal AI literature.
“The Supervised Attorney” is written for practicing attorneys, legal technologists, ethics scholars, bar regulators, and policymakers who must navigate the intersection of machine intelligence and professional responsibility. It is not a technology primer for the uninitiated, nor a treatise for AI engineers.
It is a practical and scholarly examination of what it means to use AI responsibly in a profession whose central obligation, competent representation of a human client, cannot be delegated to a model that has never passed a bar exam.
The book examines how current AI systems process legal text, where they produce reliable outputs, and where they fail against the demands of formal legal reasoning. It addresses professional responsibility directly, analysing attorney obligations under Model Rules when AI tools generate legal work product, and what supervision means in practice when the supervised party is a machine.
It also covers liability allocation when AI-generated legal advice causes harm, the tension between black-box model performance and the constitutional requirement of reasoned adjudication, and how training data reflecting historical disparities in charging and sentencing propagates into predictive legal tools.
A dedicated section of the book draws on Snomi’s operational experience building VerdictAI. It covers AI system design for high-volume, process-driven legal practices, including intake scoring models, medical record analysis pipelines for personal injury cases, and settlement negotiation support systems. This section fills a gap that existing legal AI scholarship has largely left open.
“The legal profession is undergoing a structural transformation,” Snomi said.
“Artificial intelligence is no longer a distant promise. It is a present-day instrument being used in courtrooms, law offices, and intake centers across the world. The question is no longer whether lawyers will use AI. It is whether they will use it with the rigor and accountability the profession demands.”
The book also engages with the broader regulatory landscape, comparing AI governance approaches across the United States, European Union, and United Kingdom, and examining what third-party auditing of legal AI systems should require and why voluntary frameworks have so far fallen short.
“The Supervised Attorney” is published by Lambert Academic Publishers and is available now on Amazon and other major e-commerce platforms.
Join BusinessDay whatsapp Channel, to stay up to date
Open In Whatsapp
