In early 2025, Microsoft CEO Satya Nadella reignited interest in a 160-year-old economic principle known as Jevons paradox, citing it in the context of artificial intelligence’s (AI’s) rapid evolution. His remarks, which quickly went viral, highlighted a counterintuitive truth: increased efficiency often leads to increased consumption. While this principle has long been recognized in energy economics, its implications for the legal profession are only now coming into focus.
First described by British economist William Stanley Jevons in 1865, the paradox emerged from his study of coal consumption. Jevons observed that as coal use in steam engines and industrial applications became more efficient, coal consumption increased. The reason? Efficiency made coal-powered energy cheaper and more attractive, leading to increased and broader use.
This principle has since been observed in many domains—fuel-efficient cars, cheaper data storage, faster internet—and now, in the legal profession, where AI and automation are bringing transformations to the legal services arena. As AI boosts efficiency in legal services, Jevons paradox suggests this will expand—not reduce—the demand for legal work, thereby transforming the profession and at the same time raising interesting ethical questions.
The legal field is undergoing a technological renaissance. AI tools promise to speed up processes, reduce costs and make legal services more accessible. While some are concerned that broader AI adoption may lead to the loss of law firm jobs, Jevons paradox suggests AI driven gains in efficiency may not reduce the overall demand for legal work—and in fact, may do the opposite.
Here are several key areas where this paradox has the potential to play out.
Finding case precedents, examining statutes and regulations and analyzing relevant legal issues are fundamental tasks that are essential to the core work done at every law firm. Generative AI (GenAI) tools like CoCounsel and Lexis+ AI can reduce research time dramatically, as well as the time that is necessary for reviewing and organizing that research. Updating case citations as well as case reviews and summaries can be completed in minutes, not hours. Will faster research mean some clients may expect more thorough analysis, and will lawyers be expected to explore more angles and precedents than previous budgets may have allowed?
While the potential impacts of AI are not always clear, document summarization is a use case that many lawyers can easily see and understand. Indeed, it is likely one of the more common applications of AI that lawyers or law firm staff have tried. AI is very good at summarizing the contents of one or more long documents. Some AI tools can also create chronologies from a collection of documents. While junior lawyers are probably happy to have a tool that helps them more quickly summarize documents, the ability to more quickly review and summarize documents could help firms take on matters they might not otherwise consider.
GenAI enables rapid document creation. First drafts of contracts can be created in seconds. They won’t be perfect and will have to be reviewed, but they will be very good first drafts. And existing contracts can be analyzed, formatted and created in far less time than traditional manual drafting methods allow. Will the greater ease of drafting lead to a proliferation of contracts, each requiring negotiation, review and compliance tracking? And will the document review capabilities of AI make managing large numbers of contracts much easier?
AI chatbots and legal self-help tools can empower individuals to recognize and pursue legal remedies they might not otherwise have pursued or even known about. While many may represent themselves, their initial research or steps they take on their own may give them confidence or need to contact and engage a lawyer. This expanded legal work increases filings and legal activity, especially in small claims, housing, family and immigration law.
E-discovery is another area where AI tools are already more commonly used. Not only do they accelerate document review, but they can also complete this task more accurately than humans can. Working together, humans with AI support can complete e-discovery better, faster and cheaper than humans or AI can do on their own. As these tools get better and become more widely used, will they encourage broader discovery strategies and more complex litigation? Will they help a small firm take on a litigation file that only a larger firm would have the resources to do with a traditional human-review approach?
Regulatory compliance is a complex and specialized area of law. Automated compliance tools make it easier for businesses to check their regulatory compliance and to identify risks. AI tools can complete these tasks far more quickly, accurately and cheaper than humans can. Will this open the door for smaller businesses to do regulatory and risk reviews that they would not have undertaken due to the cost?
IP law is another specialized and very technical area of practice. AI-assisted IP tools lower the barriers to filing patents and trademarks, and for monitoring registrations and their expiry dates. Will this increase the volume of IP-related legal work?
Online alternative dispute resolution (ADR) platforms make dispute resolution more accessible but also increase the number of disputes being formally addressed. This may reduce litigation, but it opens the door for more people who might have skipped the courts due to the high cost and to seek ADR.
New powerful predictive tools can analyze possible litigation or commercial dispute outcomes. This will encourage and support more strategic litigation, potentially increasing the number of cases pursued.
As AI-driven efficiency reshapes the legal profession, it also raises new and profound ethical questions. The paradox of increased demand must be balanced against the profession’s core values: justice, fairness and integrity.
With AI enabling faster output, there’s a risk of prioritizing volume over quality. Lawyers must ensure that speed does not compromise the depth of legal reasoning or the accuracy of advice.
Clients may not fully understand when AI is being used in their legal matters. Ethical practice requires transparency about the role of automation and the limitations of AI-generated outputs. Transparency in billing also helps clients understand the value of these tools and the impact on their total fee.
AI systems trained on historical legal data may perpetuate existing biases. This is especially concerning in areas like criminal justice, housing and employment law. Legal professionals must scrutinize AI tools for fairness and overcome any potential bias, especially in research results.
When AI tools make errors—such as misinterpreting a statute, hallucinating case law or generating flawed contracts—who is responsible? Like any work completed by paralegals or junior associations––and any other technology tool–– lawyers must review output with a critical eye and remain accountable for the outputs of the tools they use, even if those tools are highly automated.
There’s a danger in deferring too much to AI, especially in nuanced areas of law that require human judgment, empathy and ethical reasoning. The profession must resist the temptation to automate away its most human elements. It is this human element that makes lawyers so integral to the provision of legal services.
Jevons paradox reminds us that efficiency is not the same as reduction. In the legal profession, AI and automation are not replacing lawyers—they’re expanding the scope of what lawyers can do. As legal services become more efficient, they also become more pervasive, more complex and more in demand.
Solo and small firms are well-positioned to take advantage of this economic principle. They are more nimble than larger firms so they can make changes to their technology, business processes and business model more quickly. They can offer a new type of AI-driven service that is both affordable to the latent market while still profitable for the firm. For example, a full-service family law firm may now also be able to deliver AI-generated/lawyer reviewed martial settlement agreements for a fraction of the former price to clients who might not otherwise have been able to afford any legal representation.
To prepare for this future, lawyers should embrace AI technology. We must learn it and incorporate it into our current work to be able to use it to expand our services to others needing affordable services. We should not be afraid of it or be afraid it will replace lawyers or the staff who work in law firms. The challenge for the legal community is to manage this growth responsibly. Efficiency should not come at the cost of equity, ethics or access. As we embrace the tools of the future, we must also prepare for the paradoxes they bring.
