Just a year and a half ago, Thomson Reuters published a report titled, ChatGPT and Generative AI within Law Firms. The report found that 82% of law firm respondents believed that while generative AI can be used for legal work, only 51% said that it should.
However, AI adoption in law firms skyrocketed from 19% in 2023 to 79% in 2024. The legal tech industry is also expected to triple in size by 2036, with a projected valuation of $65.51 billion.
These numbers are staggering and indicate a clear shift—even attorneys, traditionally cautious about industry changes, are now embracing AI as a critical tool for the future of legal practice.
As technology advances, the pace at which it’s adopted grows, too. AI-powered legal tech tools are slowly becoming commonplace in law firms and will soon be as mainstream as Microsoft Word and Zoom. We've also just borne witness to the first AI drafted complaint filed in federal court, thanks to OpenAI's o1 model, released in September 2024. While there has been some debate and scrutiny regarding the quality of the complaint, this still marks a pivotal moment in the integration of AI into the legal profession (view the full complaint here).
Joshua Dupuy, lawyer and recognized expert on the intersection of artificial intelligence and law, says:
“Navigating this transformed legal landscape will be imperative for lawyers aiming to complement AI's capabilities while safeguarding their indispensable human touch. Mastering AI tools may soon become as crucial as interpreting the law itself.”
Joshua Dupuy’s thoughts on neuro-symbolic AI in legal tech
Dupuy provides a compelling perspective on the future of law and legal tech, highlighting the technical advancements in AI we can anticipate in the years ahead.
He claims that the future of law will be neuro-symbolic, an approach combining two distinct types of AI: neural networks and symbolic reasoning. According to Dupuy, neuro-symbolic AI is going to “redefine legal research and analysis, offering unparalleled precision and depth.”
So, what does this mean exactly?
Neural networks include large language models and machine learning, which comprises many of the AI tools on the market today, including ChatGPT. This type of AI recognizes patterns and processes vast amounts of unstructured data, such as identifying trends in legal case law.
Symbolic reasoning, on the other hand, is a type of AI that uses explicit rules and logic to handle structured, rules-based tasks, such as applying statutory regulations. While some legal tech tools on the market use symbolic reasoning, they are less common than those using neural networks.
Dupuy anticipates that the tools destined to become mainstream will harness a combination of these AI methodologies, creating highly powerful and effective systems. They’ll be able to understand and interpret nuanced legal texts while ensuring compliance with formal legal rules.
For instance, a neural-symbolic legal research tool would use neural networks to analyze vast amounts of case law, contracts, and legal commentary to identify patterns and potential precedents. Meanwhile, the symbolic component would apply formal legal rules to ensure results align with statutes and procedural norms.
Unlike current AI software on the market, these tools would not only find relevant cases but also reason through their applicability, providing actionable, legally sound insights.
This doesn’t mean that AI is going to replace lawyers, but it does mean that attorneys' roles are going to shift. Lawyers will need to develop new skills and ways of working, both of which will most likely be heavily reliant on and influenced by neuro-symbolic AI.
“AI won’t just speed up the legal system — it will revolutionise it."
Renowned expert on the impact of AI and President of the Society for Computers and Law, Professor Richard Susskind, recently published an article in The UK Times titled, AI Won’t Just Speed Up the Legal System — It Will Revolutionize It. In the article, Susskind shares his insights on how AI is poised to transform the legal industry in the coming decade.
He says that AI tools for lawyers, including ChatGPT, which can already perform tasks such as legal analysis and contract creation, are just the beginning. Susskind emphasizes that AI will only continue getting smarter and essential in the legal industry and that short-term thinking is the gravest danger facing lawyers.
The future of legal service will not just be an improved version of today's legal working practices, but will create "an entirely new era of the AI-empowered client." Susskind emphasizes that AI will contribute to a more democratized legal system.
"In law, as elsewhere, the revolutionary impact of AI will not be in sustaining 20th-century providers, but in enabling citizens and organisations to undertake complex tasks without relying directly on human experts.”
For instance, AI could empower individuals to navigate legal systems independently, reducing the need for traditional legal representation in straightforward cases. Susskind also stresses that AI will require lawyers to adopt new skills, such as managing AI systems and interpreting AI-generated insights.
He concludes the article with a powerful statement:
“In years to come, the principal role of AI in law will not be to enhance today’s largely unaffordable legal and justice systems. It will be to place the law in the hands of everyone.”
But OpenAI passed the bar exam. Is this evidence that lawyers are no longer necessary?
Definitely not.
For some background, OpenAI’s GPT-4 made headlines last year for allegedly scoring in the 90th percentile of the July 2022 Uniform Bar Examination. But according to a recent MIT study, our AI-powered friend didn’t really pass with flying colors.
The study reveals that GPT-4, in fact, barely passed. Researchers suggest that OpenAI’s evaluation methods may have overstated the model’s capabilities. Their findings show GPT-4’s performance was near the minimum passing threshold, raising concerns about its ability to reliably handle nuanced legal tasks.
While GPT-4 and similar models hold promise to do a whole lot, their limitations must be carefully considered in professional applications like law. And even if GPT-4 had truly excelled at the bar, it still wouldn’t replace lawyers. Passing an exam demonstrates knowledge of legal principles, but practicing law still requires human skills that machines cannot replicate, such as critical thinking, ethical judgment, client advocacy, strategic planning, and an understanding of the human and emotional aspects of legal disputes.
A future of frictionless justice
AI's integration into the legal profession has sparked fears of replacement, but these concerns miss the bigger picture. Law is far more than research and document review: it's advocacy, strategy, negotiation, and a deep understanding of human complexities. While AI is transforming how lawyers work, it complements rather than replaces their expertise. The core of practicing law lies in nuanced judgment, ethical considerations, and the ability to build trust with clients.
Both Susskind and Dupuy offer detailed and intriguing analyses on what the future of law will look like: Susskind envisions a legal world with even more advanced AI tools than we have now. He foresees a legal industry where advanced AI tools empower individuals to navigate legal systems more easily, democratizing access to justice. Dupuy predicts that the future of legal tech will be heavily influenced by neuro-symbolic AI, making the industry more precise and adaptable.
What ties these perspectives together is the idea that technology isn't replacing lawyers—it's enabling them to do more, better. Automating complex but repetitive processes allows AI to accelerate so many legal processes. This will include a reduction in legal costs, lowering barriers for individuals seeking legal help. Tools that offer self-service legal support have the potential to bridge the justice gap, making legal services more accessible, transparent, and affordable for all, not just those who can afford it.
At Darrow, this vision of a more inclusive legal system drives everything we do. AI isn’t just a tool for efficiency, but it's the foundation for creating a world where justice is truly within reach for everyone.
Evyatar Ben Artzi, Darrow co-founder and CEO, explains that the traditional methodology for detecting legal violations is slow, scattered, and oftentimes, completely inefficient. But AI and technology are changing this and allowing governments and attorneys to detect wrongdoings more easily and for less money.
He explains:
“[AI] could help energize public enforcement through government authorities, increasing their coverage. And when that’s not enough, AI could augment private law firms fighting the good fight to file lawsuits against wrongdoers. If law firms and government agencies start using AI, the court system will have to adapt, cutting through the noise and shortening durations of court proceedings.
Of course, this won’t happen overnight. But it WILL happen 一 restoring the justice system to its role as the immune system of our markets and industries. When that happens, AI might become less catastrophic and more useful at protecting humans, strengthening our societies and institutions.
Imagine it: nurturing artificial intelligence with a profoundly human intelligence. A world in which technology helps us uphold our human values and protect our rights. It exists. Let’s bring it here.”
_______________________
This might interest you:
Partner with Darrow to grow your practice
