eDiscovery, legal research and legal memo creation - ready to be sent to your counterparty? Get it done in a heartbeat with AI. (Get started for free)
The origins of legal AI can be traced back to the 1980s, when researchers began exploring how computers could be programmed to mimic human legal reasoning. Some of the earliest successful legal AI systems focused on modeling statutory analysis and legal documents. For example, researchers at Carnegie Mellon University developed a system called HYPO in the late 1970s that could analyze hypothetical fact scenarios and make legal arguments using a database of case law. In the 1980s, MIT researchers developed a system called TAXMAN for analyzing tax regulations.
These early expert systems demonstrated AI's potential in law, but were limited by the computational power and training data available at the time. It wasn't until the 2000s that legal AI started to take off in commercial settings. With the rise of big data and machine learning, legal tech companies began applying AI to automate routine legal work. The earliest commercial applications were in electronic discovery and contract review. Startups like Blackstone Discovery pioneered the use of AI for document review in legal cases. Other companies focused on automating contract analysis and due diligence.
Over the last decade, advances in natural language processing have expanded the capabilities of legal AI. Today's systems can not only read legal documents, but also generate summaries, predict case outcomes, suggest edits, and even craft new legal arguments. AI is also powering innovations in legal research, dispute prediction, and regulatory compliance. As training datasets and computational power grow, so does AI's disruption of the legal sector.
Law firms have remained resistant to change, clinging to traditional models and manual processes even as other industries have modernized. But this status quo is on borrowed time. Several factors make law firms ripe for an AI-led shakeup.
First, legal work is drowning in data. The average case now involves tens of thousands of documents and emails to review. Firms are overwhelmed by the amount of materials they must sort through. This deluge of data creates huge inefficiencies as lawyers waste hours searching documents or miss key evidence. AI-powered tools like predictive coding and analytics can help firms automate document review and extract insights from massive datasets. Adopting legal AI can massively speed up legal work.
Second, billable hours reward inefficiency. The billable hour model incentivizes manual busywork. For decades, law firms have leveraged associates to pad invoices through brute force grunt work. But AI automates the grunt work, empowering lawyers to focus on higher-value tasks. Forward-thinking firms realize AI improves productivity and quality. Clients will demand firms work smarter, not just rack up hours.
Third, junior lawyers face punishing workloads. Legal AI can alleviate the grunt work burden on associates. Recent surveys show junior lawyers at top firms bill over 3,000 hours a year doing document review and research. Such work grinds down young lawyers, contributing to high attrition rates. Offloading rote work to AI makes junior lawyer roles more sustainable and fulfilling.
Fourth, legal research is ripe for automation. Even with today's research tools, lawyers waste hours finding cases and precedents. An AI that reads and synthesizes case law could massively boost research efficiency. Startups are already productizing legal research AI. Incumbents can either adopt this tech or get left behind.
Finally, client expectations are changing. Corporate clients increasingly view legal services as a commodity and scrutinize value. They want faster turnarounds and more certain budgets from their outside counsel. AI allows law firms to tighten timelines, reduce surprises, and quantify the value they provide. Firms that don't adopt legal AI risk losing market share to innovators that deliver smarter, faster services.
While AI still has limitations compared to human intelligence, it possesses uncanny superpowers in a few key areas that make legal work a prime target for disruption. Specifically, AI offers massive improvements in speed, scale, and consistency.
On speed, AI systems can ingest, process, and analyze legal documents magnitudes faster than any human lawyer. For example, contract review AI can parse hundreds of pages of agreements per minute and flag key terms. This enables outside counsel to provide clients incredibly fast turnarounds on due diligence for deals and transactions. An international law firm recently used contract AI to review 20,000 pages of agreements over a weekend - a task that would have taken weeks of manual effort.
For document review in litigation or investigations, AI tools powered by optical character recognition (OCR) and natural language processing can tear through gigabytes of texts, emails, and scanned images to find critical evidence and build timelines. This accelerates investigations while reducing the risk of missing "hot" documents. Lawyers need only handle the most complicated documents while AI triages the rest.
On scale, AI has virtually unlimited capacity for legal work. While even the largest law firms may employ a few thousand attorneys, cloud-based AI has no constraints. It can analyze millions of pages of documents across hundreds of matters in parallel. AI doesn't suffer from mental fatigue or burnout. Law firms can leverage this artificial scale to serve more clients simultaneously without expanding headcount.
An example is Luminance, an AI discovery tool trained on millions of legal documents. Luminance can ramp up instantly to tackle massive litigation datasets. There's no need to add more junior lawyers. This effectively makes AI associates infinitely scalable.
Finally, AI offers superior consistency compared to human analysis. Unlike people, machine learning models apply the same logic without fail every time they process an input. For instance, when reviewing contracts or case materials, AI extraction tools will accurately tag every instance of the same clause or legal concept. Humans get inconsistent when dealing with high volumes or complex documents.
Lawyers at Orrick used AI from Kira Systems to analyze commercial loan agreements 10x faster than manual review while improving accuracy. AI doesn't overlook or misinterpret terms due to fatigue or negligence. This consistency builds confidence in results and lowers errors.
Legal research is a prime target for AI automation. Hours spent searching for relevant statutes and precedents represents major inefficiency under the billable hour model. Yet automating legal research also raises concerns about potential drawbacks. Understanding the pros and cons will help firms adopt wisely.
On the promise side, applying natural language processing and neural networks to legal texts can significantly enhance search efficiency. For example, Casetext's CARA tool reads through millions of court opinions to surface the most relevant results and favorable precedents for a legal issue. It also recommends related cases lawyers may have overlooked. This automates much of the discovery process for building persuasive arguments from case law.
Meanwhile, companies like ROSS Intelligence are building advanced QA capabilities that allow lawyers to simply ask questions and get citations. Instead of combing through indexes and manually reviewing documents, the AI pulls answers directly from its vast database. It's like having instantly searchable case law encyclopedias.
Experts see particular potential around algorithmically determining the importance of cases. Since key precedents tend to be cited frequently, AI can help surface pivotal rulings and even predict their implications for open questions. Startup Ravel Law uses network analysis to map relationships and implications between cases based on citation patterns. This provides lawyers invaluable insight into how a court may rule on novel issues.
However, concerns exist around potential for bias, errors, and misplaced reliance. Some lawyers worry algorithms trained on past precedent could overlook outlier cases that don't fit historical patterns but are vitally important. Outlier cases often change legal interpretation. Wholly relying on AI could cause lawyers to miss creative angles.
There are also fears about encoded bias. Since datasets reflect society's existing biases, algorithms can entrench unfairness in application of law. Critics argue legal research AI needs more human perspective to avoid just reinforcing status quo assumptions.
Finally, lawyers may become over-reliant on automation. If human expertise at searching case law atrophies from disuse, it could leave firms dependent. People may lose capacity to critically evaluate results. And poor data hygiene in training sets means algorithms will inevitably serve up some false positives.AI should augment, not replace, human legal research skills.
The rise of AI begs the question - could machines fully replace junior lawyers and paralegals? Or will AI simply transform their roles? Many see the writing on the wall. A recent Deloitte survey found 51% of law firm partners expect AI will reduce lawyer headcount in the next 5 years. Yet views diverge on the extent of displacement versus augmentation.
A common perspective is that routine work like document review will shift to AI while uniquely human skills remain. As Dechert LLP's CTO remarked, "There are certain tasks that are ideal for machine learning algorithms, but that doesn't mean the entire profession is at risk." Bloomberg Law"s CEO sees AI as "allowing humans to do the things that lawyers went to law school for " critical thinking, persuasive arguments."
This view resonates with junior lawyers feeling burnt out from grunt work. In one survey, nearly half of firms said AI could prevent associate attrition by eliminating drudgery. Partners envision AI as a lever to banish grunt work entirely from the associate role to focus purely on high-value analysis.
Yet critics argue viewing associates as just future partners undervalues their present contributions and humanity. "We have to get away from treating associates like machines," urges law professor Bernard Burk. This view holds AI should transform law rather than entrench the status quo.
Some even see full replacement as likely. Richard and Daniel Susskind, widely seen as legal futurists, declare "the end of lawyers" in a provocatively titled book. They argue legal roles will splinter with AI performing the bulk of core tasks. A partner at White & Case predicts paralegals and associates will be 90% displaced within 15 years.
Many lawyers push back against the notion that AI could replace them. Their counterargument is that legal work requires human skills and judgment that software cannot replicate. Specifically, they emphasize irreplaceable human abilities like critical thinking, creativity, empathy, and counseling.
A core objection is that lawyers don't just research rules but craft novel legal arguments. As Stanford law professor Pamela Samuelson puts it, "Computers are good at crunching through documents and identifying relevant passages that have particular key words, but they"re not going to come up with creative new legal arguments that haven"t been made before." Even AI tools that summarize case law cannot conceiveoriginal theories to interpret that law. Human ingenuity remains indispensable.
Relatedly, lawyers see legal judgment as too subjective for automation. Factors like witness credibility, judicial leanings, and cultural context shape how lawyers apply the law. An algorithm cannot assess how sympathetic a judge will find a plaintiff. Assessing multiple murky factors and making tough judgment calls is innate human expertise.
Many also emphasize the soft skills involved in practicing law. A bot cannot establish empathy and trust with clients during counseling. Negotiation and persuasion rely on emotional intelligence machines lack. As attorney Miranda Jones remarked, "How do you code instinct? There are intangibles that humans bring to bear." The human touch will continue distinguishing top lawyers.
Meanwhile, lawyers believe their qualitative skills offset any efficiency gains by AI. Legal ethics require carefully evaluating conclusions from any analysis. While AI may research faster, humans must validate results rather than blindly relying on technology. human oversight is essential.
Overall, skeptics feel portraying the legal profession as easily automatable overlooks its human complexities. They argue AI will remain only an accessory because the essence of lawyering revolves around creativity, judgment, persuasion and counseling. These innately human abilities mean lawyers will stay irreplaceable.
In recent years, eDiscovery and due diligence have emerged as prime use cases demonstrating AI's transformative impact on legal work. These labor-intensive processes epitomize the document-heavy grunt work associates slog through. AI has proven it can automate substantial parts of this work and augment human capabilities. Law firms that fail to adopt AI for discovery and due diligence risk falling behind.
A permanent shift occurred when Blackstone Discovery broke new ground applying AI to legal document review in the 2000s. The company used technologies like concept searching, predictive coding, and analytics to rapidly classify documents for relevance. This slashed the time and costs of discovery while also improving accuracy. Blackstone could handle massive litigation datasets on tight deadlines at a fraction of the humans required.
Seeing this success, white-shoe law firms moved to integrate similar AI tools into their process. As Shook, Hardy & Bacon partner Patrick Oot noted, AI document review "allows our firm to handle the huge volumes of electronically stored information clients now face." Partners rave about time and cost savings from AI.
The gains can be dramatic. DLA Piper saw document review efficiency jump 70% using tools like Luminance. Another firm used Kira Systems to analyze loan agreements 10x faster than manual approach. This allows delivering better service to clients. As UnitedLex's CEO observed, AI systems like Luminance achieve "greater speed, accuracy, and cost predictability" for evidence finding.
Due diligence and transaction support have also benefited hugely from AI efficiency gains. When reviewing agreements for M&A, financing, IPOs, etc. AI can provide lightning fast turnaround. For example, lawyers at Davis Polk used Kira's machine learning to evaluate material contracts for a stock purchase 3x faster than traditional manual review.
This accelerated due diligence helps lawyers meet compressed deal timelines and promptly flag key risks. The magic happens when combining humans and AI. As Davis Polk partner Oliver Cohen remarked, Kira Systems "has allowed our lawyers to focus on the qualitative, subjective, and strategic assessments".
AI also makes juniors feel less like cogs in a wheel by eliminating grunt work. As a Latham & Watkins associate noted, AI document review "lessens the burden on junior associates" and lets them focus on higher-value tasks. Partners consistently report higher associate satisfaction after AI adoption.
The future of the legal profession will be defined by effective collaboration between humans and AI. Rather than replacing lawyers, AI will augment human capabilities to create better outcomes for clients. The key for law firms will be seamlessly combining unique human strengths with machine superpowers.
Many experts advocate focus on the hybrid model. Stanford Law professor Daniel Martin Katz states that AI should enhance what lawyers do best: "The robot will do what it does best, which is crank through data and identify patterns, and the human will interface with the client and do the creative work and translate that data into strategizing."
Likewise, Erin Ellis of technology company Onna asserts that AI systems "will work alongside human team members to augment and maximize expertise, insight, and judgment." She adds, "The beauty of these technologies is that they adapt to the way people work." Collaboration is central.
Forward-thinking firms are already benefitting from integrating human and AI capabilities. International firm Eversheds Sutherland developed its own legal AI called Kira to boost productivity. As Eversheds" Chief Innovation Officer put it, "With the combination of Kira and Eversheds Sutherland lawyers working together, we believe we can offer our clients a faster, better service."
UnitedLex also champions hybrid models, with CEO Dan Reed stating: "Technology solutions today should bring together the strengths of both machines and lawyers to deliver faster and better legal outcomes." UnitedLex builds custom AI to mesh with specific client needs and workflows.
Other firms praise AI's ability to offload repetitive tasks and let humans focus on judgment-intensive work. Davis Wright partner Greg McPolin observes, "The AI piece allows me to spend more time on strategy and negotiations by letting the robots do the more mundane work."
To fully realize the hybrid model, firms must avoid treating AI as just an add-on tool. Jennifer Taylor Hodges, Chief Legal Operations Officer at DailyPay, urges truly rethinking roles: "Rather than just replacing the tasks humans used to do, it's thinking about how to redesign jobs around capabilities of people and technology."