How AI is Laying Down the Law: Past and Present
- Madison Shanfeld
- 51 minutes ago
- 6 min read
Welcome to the first installation in The Political Pen’s newest duology: How AI is Laying Down the Law. In this installment, we will be exploring how AI has penetrated the field of law and how it is currently being utilized. Included is an expert interview with D. Allan Asbury, Deputy Director/Senior Counsel-Ohio Board of Professional Conduct at The Ohio Supreme Court. The next article discusses the forefront of development for Legal AI with expert Nate Jedinak, Senior Director of Software, Data, and Innovation at Vorys.
AI's Movement into Law
Commonly accused of stealing jobs, artificial intelligence (AI) is sneaking into almost every field of work in the United States. While law is typically considered one of the careers less likely to be completely consumed by AI, it is not resistant to artificial intelligence tools becoming common in the legal workplace.
AI’s first introduction to the sphere of law was in 1961 with primitive versions of databases used to facilitate access to case law and legal documents. In the 1970s, the
program JURIS was created and served as a pioneer in the realm of computerized research. With JURIS, attorneys could sift through state and federal law more efficiently than the previous method of shuffling through thick books and legal reports in dusty libraries.
As AI has continued to develop, other pre-existing research systems began adopting AI systems to utilize natural language in their searching mechanics, making legal research more efficient and intuitive for lawyers. Instead of only understanding key words, they could now show accurate results to inquiries. Additionally, these systems began to utilize artificial intelligence that extracts key clauses and important data from long court documents, inspiring lawyers to move further away from traditional methods to data-driven predictive analysis and AI assisted research.
In the early 2000s, Stanford University professors began a side project called Lex
Machina. It marked a new development in AI as part of the rise in legal analytic platforms. Programs similar to Lex Machina would use court data on specific judge's litigation trends and behavior in court to predict the probability of success in various legal situations. It provided lawyers with historic data that would allow them to make more informed decisions than was previously possible.
Moving towards AI greatly changed the fundamental practice of legal research. Instead of having a primary focus on simple case law, it shifted to preparation of strategic legal intelligence. Additionally, analytical resources were now obtainable for all, evening the playing-field for smaller firms.
Modern Day
One of the most recent forms of AI penetrating the field of law is generative AI, or GenAI. Examples of generative AI would include systems like ChatGPT, Microsoft Copilot, or Gemini. There have already been law specific adaptations of GenAI including HarveyAI, CoCounsel, Lexis+AI, and Westlaw Precision, all claiming to assist with generating drafts, research, and alternative legal tasks.
Many law firms across the nation are increasing their use of AI in daily practice. This includes firms like Orrick, Herrington & Sutcliffe and Fisher Phillips, who have around 1,650
lawyers combined and are both utilizing Copilot. A multitude of firms across the nation are transitioning to using HarveyAI, while global law firm Baker McKenzie created a new position solely dedicated to combining machine learning, AI, and the firm’s expertise to build the optimal service. Furthermore, in 2023, Dentons and Addleshaw Goddard announced they were going to craft their own internal GenAI based off of GPT technology. Even start-up law firms are applying AI to build litigation techniques. In the modern era of law, AI is less of a tool and more of a coworker that attorneys need to be able to properly utilize.
AI's Hallucination
One of the most disturbing aspects within the legal uses of AI today is the abundance of hallucinations created by GenAI. When attorneys are asking AI systems to look for related case law and precedent for a new case or brief, there has been a trend of AI fabricating legal case citations or reporting false details to pre-existing cases.
While it is ethically improper to be citing fabricated or false precedent in a court of law, it can be construed as a violation of rules of professional conduct. Allan Asbury explained that the use of AI hallucinations in court presented documents is classified as making a false statement of law to a court in violation of rules prohibiting the same.
A Warning of Over-Reliance
Asbury claims that hallucinations with AI, and AI overuse in general, is particularly detrimental to the new generations of lawyers. Within law schools, AI education is already becoming implemented, however he worries about an abundance of reliance on the technology. In particular, he expresses the sentiment that over-reliance on AI in the legal field will start to affect the quality of lawyers.
It is not expressed as a blanket statement: but instead as an expectation and a “desire for lawyers to deeply understand the law and to be deep thinkers” (Asbury). By continuously
increasing the use of AI in this field, new attorneys are not required to delve as deeply into research and think as critically as the lawyers of decades past. Instead of reading thousands of cases and building the knowledge of various precedents through hands-on experiences, AI regurgitates this information. Therefore, the repertoire of knowledge built during a lawyer's time as an associate, is being decreased by AI use. Furthermore, when GenAI does hallucinate, new lawyers are struggling to catch the error because they lack the wisdom and extensive knowledge possessed by a seasoned attorney.
Allan Asbury approaches the topic of over-reliance from the idea that lawyers eventually become the nation’s judges, and it would be most desirable for our judges to be able to fulfill their responsibilities of interpretation based on their own knowledge of the law rather than what AI has concluded.
Should judges in the United States begin turning to AI to dictate rulings, technology would be primarily referencing precedent to make its decision, thus inhibiting some of the changes that the United States could make to the nation through overturning previous decisions. For example, in the Supreme Court Case of Brown v Board of Education, there was an overturning of precedent. The decision required human interpretation and general observations of changes in society to instigate change within the United States.
Legal AI on a Global Stage
Legal AI is being utilized differently across the globe, particularly in regard to judicial decisions. In the United States, AI use is watched, recorded, and somewhat controversial.
However, China encourages the use of AI in the justice system. They have a “Smart Court Initiative" with goals of full AI integration courtrooms by 2030. The country even requires their judges to reference AI in their decision-making processes; should the judge choose to issue a decision contrary to the AI recommendation, they must explain their deviation in great detail. China is even going as far as to establish “Internet Courts” or completely digitized courtrooms with AI-powered judges to settle disputes without a physical trial. Similarly to China, Brazil is embracing Legal AI in the judicial system. Due to a substantial backlog in cases, the Brazilian government has chosen to utilize an AI enterprise model to increase the efficiency of their courts.
Will There be Legislative Action?
When asked if there would be legal measures taken to prevent AI overuse, Asbury stated that he did not expect laws to be set in place within the near future. Instead, he says that each state will regulate AI use in their own courts. While there have been attempts at legislation on the national level, such endeavors have largely failed.
However, the American Bar Association (ABA) has already established prior rules that
are applicable to AI. Particularly within the previously stated false statements that are applied
to AI hallucinations and within the existing rules on a lawyer's independent professional judgement. In further depth, the ABA dictates that an attorney must exercise their independent professional judgement, or their obligation to provide objective, candid, advice to a client. This applies to AI because lawyers, particularly young lawyers, are not relying on their knowledge, familiarity, and wisdom to give their opinion. By utilizing AI to form and deliver their opinion they are referencing a third-party resource, possibly violating their independent professional judgement.
As Asbury says, “with the ABA, the rule structure is already there.” As of February, 2025, the rule structure within the American Bar Association and the independent rules of the State Courts are enough regulation for modern AI. However, as AI continues to become more advanced, that could be susceptible to change.
Final Thoughts
From early expert systems to modern generative models, AI has transformed how the law is researched, practiced, and understood. Yet the core purpose of the legal system, fairness, transparency, and human responsibility, has not changed. AI will continue to evolve, and in the next installment of this duology, we will explore the systems on the forefront of Legal AI's future.