It is no secret that the law office has not been a traditional pioneer on the frontiers of technology. However, recent dizzying advances in computing capability and ease of use and accompanying drops in cost are breaking these long-standing barriers. Law practices are embracing technology in all its aspects, from everyday tech to improve communication and practice management to advanced data tools such as artificial intelligence (AI) to increase efficiency and quality of services provided to their clients. In spite of, and perhaps partly as a result of, the Covid-19 crisis and ensuing shift in business patterns, industry investment in AI technologies is expected to accelerate and increase five-fold over the next few years, reaching nearly $200B globally by 2025.  

Advanced analytics, including the use of AI and machine learning, stand poised to advance the practice of law by enabling cheaper, more consistent, and faster decision making and overall a nimbler and more competitive professional practice. While the legal field has employed AI for some  tasks (such as e-discovery support and research) for over a decade, the use of AI is now expanding into new areas of practice, from easing regulatory complexity to hiring decisions, contract analysis, and even legal document generation. These modern practice methods bring several new considerations alongside their benefits, from a need for increased accountability and transparency to discrimination, data privacy, and security concerns.

At its core, AI is a set of computing processes that automates and improves processing of large data sets. It functions by analyzing the data according to predetermined of rules and revealing patterns, leading to new insights and better, faster answers and conclusions for its users. One application of AI, machine learning, not only applies predetermined rules but also learns from discovered patterns and self-develops new rules to apply to future tasks and ultimately better predict outcomes. Other AI applications include natural language processing, or the processing of words in the context in which they appear, part of the bedrock of law practice and commonly used in e-discovery platforms such as Relativity, Everlaw, and Catalyst and legal research tools such as LexisNexis, Westlaw, and Casetext. Beyond its long-standing uses, AI is being introduced to other areas of law where the impact can be just as profound: regulatory reform and contract review and analytics. 

At the federal level, the Regulatory Flexibility Act at 5 U.S.C. § 601 was enacted in 1980 to establish and maintain rules and regulations that fit the purpose and scale of the businesses, organizations, and jurisdictions to which they apply.  It mandated that agencies “plan for the periodic review of rules which have or will have a significant economic impact” and ensure that they are regularly adjusted and tailored to their purpose or ended altogether when their usefulness has passed. This need is evident given federal rule complexity, overlap, conflict, and duplication, as well as corresponding changes in society, technology, economy, and business practices over time. With the Act in mind, the Office of Management and Budget (OMB) tackled the fearsome Code of Federal Regulations in a 2019 wide-reaching review. It used an industry-developed process and tool, RegExplorer, which combines artificial intelligence and subject matter experts to analyze, map, and identify trends and hot spots among rules that can be used to prioritize review and attention by regulators.  Some examples of their findings include hundreds of broken citations and references and rules that on average have not seen changes in nearly twenty years. Further, more than two thirds of the CFR sections have never been edited and nearly ten percent of CFR sections are “extremely close matches” to others which indicates substantial overlap with the potential for conflict.  This tool has also been used at the state level, with Ohio’s Common Sense Initiative using RegExplorer in 2020 to monitor its state rules not just for duplicates but also to hopefully “inform the drafting of legislation” based on better context and applicability. Other countries including Canada and Australia have announced interest and similar initiatives, indicating a clear appetite for using advanced data analytics and AI to study and resolve the often complex, politically fraught questions of regulatory reform.

Contract review, analysis, and compliance management is another recent advance in which AI and advanced data processes are supporting law offices and in-house counsel.  The orientation of the world’s economies towards ever more integration and global reach has driven business interests towards more complex arrangements.  The resulting intricate contracts and frequently accompanying multitude of service agreements and licensing documents would soon rival phonebooks in size if we were not in a digital world.  Understanding contract provisions and their effects, often dependent on law and rule interpretation and references, but also rife with standardized phrases and language is a prime target for technology assists including AI.  The combination of standard approaches of natural language processing augmented with customized and ever evolving rules of machine learning puts AI in a prime spot to drive improved contract practice.  Better analysis and recognition of omissions, conflicts, and weaknesses will result in more effective documents, less costly litigation, and less conflict for businesses.  Contract-focused AI tools can offer many benefits such as putting contracts in the context of their industry - for example, ensuring appropriate compliance language is used for financial transactions. Another use might be putting contracts in the context of their organization, analyzing potentially thousands of company contracts simultaneously to identify outliers or issues, out of date provisions, or inconsistencies in language, again allowing for the optimization and prioritization of limited legal time and resources.  The tools and companies working in this space are numerous and include Kira, Ross Intelligence, and Luminance with many getting their start not from big law offices but in University mathematics, engineering, and science departments which drive AI and computing developments to new heights.

The potential disruptive power of AI and machine learning and its notable impact on regulatory frameworks, legal practice, and business operations has not escaped the eye of regulators.  In 2019, we saw the introduction of the Algorithmic Accountability Act which required “entities that use, store or share personal information to conduct automated decision system impact assessments and data protection impact assessments.” While the act did not ultimately pass, the OMB issued in November of 2020 final Guidance for Regulation of Artificial Intelligence Applications which establishes a broad framework for agencies to assess potential regulatory and operational approaches to emerging AI uses and  issues.  This stimulates development in the field while also enabling and encouraging regulations of AI technologies.  Alongside US progress, the European Union has passed rules on robotics and ethical guidelines for commercial AI solutions.

The use of AI and machine learning is already omnipresent and growing, powering vast improvements in our current data age. It is in our homes through virtual assistants and smart devices. It powers self-driving cars, our web searches, utility grids, business, and healthcare; and it’s making continuous strides into law practice as well.  We may be on the precipice of a bold new world powered by AI and we should consider its advantages, shortcomings, and blind spots, and harness its power through appropriate regulations while also maximizing its benefits to business and society.

Andrei Dumitrescu is the Chief Data Officer and Deputy General Counsel at the State of Tennessee, Division of TennCare. In this role he is part data strategist and adviser, part counsel to ensure regulatory compliance, and driver of efforts to maximize the quality and programmatic value derived from agency data assets.  TennCare is Tennessee's Medicaid program which provides health insurance coverage to 1.5 million Tennesseans in certain groups including pregnant women, children and caretakers, older adults, and those with disabilities.