AI legal robotsgettyAs artificial intelligence reshapes many industries, the legal field faces its own crossroads. Over the past few years, a growing number of legal professionals have embraced AI tools to boost efficiency and reduce costs. According to recent figures, nearly 73% of legal experts now plan to incorporate AI into their daily operations. 65% of law firms agree that "effective use of generative AI will separate the successful and unsuccessful law firms in the next five years."Investors have shown strong support for AI-powered legal startups, with funding reaching new record highs in 2024 with total capital investment of $477 million. The appeal for VCs is the potential that 44% of legal work could potentially be automated by emerging AI tools. Startups like Harvey, raised a $100 million Series C round at a $1.5 billion valuation.We explore the current state of legal automation with Ben Su, Co-founder and Head of Growth of Capita, the worlds first AI lawyer, and discuss how this shift towards AI could reshape the delivery of legal services. Carey Lening, a legal-tech consultant and recovering attorney who focuses on privacy and data protection and Jide Afolabi, a probate lawyer, a graduate of Osgoode Law School with over 20 years of experience weigh in on the promises and pitfalls of automation in the legal sector. What are the challenges and criticism facing this change? What are its effects on legal education? What will AI might mean for the future of work?The State Of Legal Automation TodayAI has entered the legal sector with some noticeable force. The billion-dollar valuations experienced by companies like Harvey has created an investment frenzy that saw 58 funded deals in 2024. AI automation to speed up contract analysis, document summarization, and case research means a shift away from the slow, traditional methods that have defined the legal industry for decades, saving 4 hours per week and the opportunity to increase annual billable time per lawyer by $100,000.Lening, who has 20 years of experience in technology and data protection, disputes the slow pace of technological adoption in the legal sector, Back in 2001, when I was working at a law firm, we were already using WordPerfectnot just for basic text processing, but specifically because it supported macros and structured templates. These tools allowed us to automate repetitive tasks and standardize documents, even then. In reality, automation has been part of the legal field for quite some time.What has changed is not the existence of automation in the legal field, but rather the sophistication of automation tools now available, as Lening adds, The industry has evolved far beyond basic macros and Excel spreadsheets. Todays technology offers entirely new opportunities to automate legal tasks in ways that were previously impossible.Afolabi, who holds a masters from the London School of Economics, describes the evolution of legal processes over the past five years, highlighting the shift from paper-based systems to automated ones. He explains that the initial client interaction, where they tell a story and paint a picture remains crucial. However, the method of capturing and analyzing this information has changed significantly."Five years ago, that would have been done via paper. Youre taking notes," Afolabi states, "now, there's automation for that." He emphasizes that while the core process of asking questions remains, it's now "the machine asking the questions." Automation extends to the initial risk analysis, where the system can contextualize the kind of issues and how to best proceed.Afolabi stresses that this automation doesn't replace the lawyer entirely: "There's still a lawyer there with the clients, of course." Instead, it enhances the process, making the initial intake and risk analysis more efficient. He describes the document production process as "incrementally automating," but emphasizes a balanced approach: "It's a combination and I'm not entirely leaving it to the machine to do it. But I am cueing to see what the machine can produce for me."Su, of Capita, challenges this view arguing the sector is entrenched in inefficiencies, "The legal industrys obsession with making lawyers more profitable misses the point. Optimizing broken workflows isnt innovationits entrenchment. We need systems that eliminate inefficiencies, not perpetuate them." He maintains that industries like law are stuck in outdated systems designed with human-limitation constraints where "inefficiencies arent just toleratedtheyre baked into the business mod, and no one is challenging those existing models.Lening asserts there are progressive strides she and her colleagues have made to address AIs potential in the field such as Leonard Park and Dazza Greenwoods use of LLMs for contract analysis, and Michael Bommeritos Kelvin, specialized model trained to handle various legal and financial documents, including SEC filings. She also shares her own experiments:"I wrote a Python script with ChatGPT to summarize cases. It worksbut it took quite a bit of iterationboth of the code itself and the promptsto get there. The key is knowing how to ask the right questions... I'm dubious about statistics like AI reduces contract error rates by 90%. Lawyers need to understand these tools arent magical oracles. They require precise prompting and context. If you ask an AI to find all bad clauses without defining what bad means, its useless."Ben Su, Capita, And The Worlds First AI LawyerSu went to law school but never planned on settling into a conventional legal career. Instead, he aimed to understand the rules governing business transactions and apply that knowledge to innovate within the legal system. Eventually, Su began practicing law at a boutique firm in Toronto, where he reconnected with a former classmate who shared his vision to use technology to overhaul outdated legal practices. Together, they founded Capita, a company dedicated to automating the legal processes that often slow down startup growth. Their first product streamlined the capital formation process, cutting down on the need for manual legal work and processing transactions at an impressive rate. Within just a few weeks of launch, the company was handling transactions worth thousands of dollars each day.Reimagining Legal Services With AIThe promise of AI in legal services goes beyond making lawyers work more productively. Su envisions a system where AI serves as a constant, always-on companion to clients, offering advice and identifying issues before they escalate. "Clients don't care about what tool you're using at the end of the day," Su states, "they care about the outcome and how much you're charging for that outcome." This questions the traditional hourly billing system that has made legal services highly profitablea consideration for smaller businesses and startup organizations, who often find themselves priced out of traditional legal services.In Sus model, technology handles the heavy lifting of data processing while human professionals step in only when their judgment is truly needed. Capita is already making strides. Its products ask detailed questions that mimic the initial discovery process at a law firm. The answers allow the AI to generate tailored legal documents and strategies, speeding up the process while ensuring advice fits each clients needs.Su believes that automating parts of the legal process can encourage transparent and affordable pricing. Instead of unpredictable hourly bills, clients pay a fixed monthly fee for certain services.Afolabi agrees that the traditional hourly billing model is outdated but acknowledges that many lawyers prefer this model because "it maximizes value for them." He explains that instead of reducing hours or changing the billing model, some firms might maintain their revenue by inflating billable rates: "What they might do instead of reducing the hours or changing the model entirely is to increase the hourly fee to still get the same amount of billing out of work that may be used to take 100 hours and not take 40 hours because the machine does the rest of it."What will ultimately drive costs down, he admits, is competition stating, "Like in any other industry, if a few firms in a specialized sector reduce their costs, it will pressure others to follow suit. The first firm to blink will drive the cost down." However, Afolabi notes that the specialized nature of legal services can make price competition challenging. He states, "If you have only seven firms competing for a specialized service, this, in itself, makes it easier for firms to maintain current pricing structures.Lening acknowledges Sus advocation for disrupting the current model and points out potential changes in the pricing dynamics of legal services. She suggests that clients may begin to challenge high fees, especially if they suspect that AI tools are being used to perform work traditionally done by lawyers: "There may be some interesting dynamics at play in terms of how much big law firms can charge to their clients.Facing the Challenges And Criticism of AI EfficacyDespite its potential benefits, the push toward legal automation has not been without its critics. Many traditional lawyers argue that legal work is too subtle for machines to handle. The worry is that AI systems, which largely rely on predictive text and pattern recognition, may not fully capture the complexities involved in legal agreements.About one in four legal practitioners view AI as a threat, expressing concerns that reliance on technology could diminish the human judgment needed for complex legal matters. While AI systems can process data and generate drafts at speeds far exceeding manual work, the worry is that an overreliance on automated tools might lead to errors that only a seasoned lawyer can catch. Stats show that AI tools hallucinate in at least 1 out of 6 legal queries. This tension between speed and careful judgment continues to spark debate among those who work in the legal field.Lening underscores these risks, particularly for non-lawyers: "Consumer-facing legal AI scares me. People lack the smell test to catch nonsense. Look at Joshua Browders DoNotPaythe FTC fined them for overpromising. Clients dont know what they dont know, and AI cant fill that gap yet." Lening refers to DoNotPay, "the worlds first AI-robot lawyer, founded by Joshua Browder in 2015. Browders application used chatbots to guide users through legal processes to fight parking tickets, cancel subscriptions, among others with the aim to make legal services more affordable. In September 2024, DoNotPay received a fine from the Federal Trade Commission for falsely advertising their AI services capabilities.Su is aware of these challenges. He notes that many founders have used generic AI-generated templates with disastrous results: "If you use ChatGPT for specialized legal work, youre guaranteeing a negative outcome. Weve fixed countless errorslike founders shutting down startups due to flawed AI-generated equity splits."Lening critiques the legal industrys reluctance to educate itself: "Many lawyers think AI is a magic box. Theyll dump a contract into ChatGPT and call it a day. Thats where disasters happen. We need better education on both the capabilities and limitations."Rethinking Legal Education And Professional DevelopmentOne of the more challenging issues raised by the shift toward AI in legal services is the role of legal education. Su shares his own experience as a law student who, despite strong academic performance, struggled to secure a position after graduation. He criticizes the traditional approach, where the burden of unpaid labor falls on those just starting their careers. I think the current system is extremely unfair, Su remarks, noting that law students are often exploited to provide free labor for law firms.He challenges the apprenticeship model: " By introducing AI into legal services, there is the potential to reduce the workload on young lawyers and offer more balanced opportunities. With AI handling routine tasks, junior legal staff might have more time to engage in meaningful work that adds real value for clients "Lening reflects on the evolution of legal education and the impact of technology and acknowledges that while the traditional law school experience, including the Socratic method, can be challenging, it serves a valuable purpose. Lening explains, "It does force you in a very consistent way to think differently about problems, to break things down into discrete things. You learn how to analogize better. You learn how to craft your thoughts in a way that becomes extremely useful and becomes a valuable skill."However, she advises prospective law students to be more strategic about their decision to pursue a legal education. She states, "I would tell people wanting to go into law school to be far more strategic about why." She emphasizes that the days of entering law school simply because one cant figure out what else to do are over, noting, "It has changed from being a thing that you can doget out ofand still be fine to something that you really do need to evaluate a little bit more because you happen to test well and you like to argue."Lening also highlights the changing landscape of the legal profession due to technological advancements. She suggests that these changes will lead to new applications of law degrees, stating, "What these tools and what technology and what the law changes in general are going to continue to manifest is that people are going to be using their law degrees in different ways, in new ways."Lening calls for updated curricula, emphasizing "Law schools teach analytical thinking, but graduates need technical literacy too. Future lawyers must learn to audit AI outputs, design prompts, and collaborate with engineers." She suggests that while significant changes are happening, many of these shifts have already been underway for some time, and observes, "I think that the existing model that we all, conceptually, have in our minds about working in a big shoe law firm and becoming partner has already started to shift." The traditional career trajectory in prestigious law firms is no longer as straightforward or guaranteed as it once was. However, she acknowledges that there will always be a market for top-tier law firms, noting, "... because people like paying for exclusivity and for what they think is the best."Broader Reflections On The Future of WorkThe debate over legal automation extends beyond the legal field. Similar questions are being asked in industries ranging from manufacturing to healthcare. While new technologies often spark fears of job loss, history shows they also create opportunitiesonline platforms, for example, have birthed entirely new professions.Lening compares AIs integration to past technological revolutions: "Were in a horse-to-car transition. Laws and norms will adapt, just as they did for automobiles. AI wont make us lazieritll create new problems to solve. Thats how progress works."She remains optimistic stating AIs role and concurs with Su, who emphasized human oversight can mitigate risks, AI isnt about eliminating humansits about enhancing efficiency. Lening added, the fourth industrial revolution isnt about job loss. Its about redirecting human talent. Let AI handle document review; let lawyers tackle creative disputes or regulatory challenges. Thats the future."Progress Means Challenging The Status QuoWhile concerns about AIs limitations persist, the consensus is clear: AI-driven services like Capita can make legal services more affordable and accessible without replacing human oversight.Lenings perspective challenges the future role of lawyers: "Automation isnt about replacing judgment. Its about freeing lawyers to focus on what humans do best: empathy, ethics, and innovation. The future isnt AI versus lawyersits lawyers with AI."