What the Past Two Years of GenAI Mean for Legal’s Future

Artificial intelligence (AI) has rapidly become a transformative force across industries, from personalized healthcare recommendations to algorithmic trading in finance. Yet perhaps no industry is as cautious, complex, and careful in its embrace of AI as the legal field. The very foundations of law are built on precedent, a legal practitioner's judgment, and a profound understanding of nuance, which can seem at odds with the algorithmic nature of AI.

The past two years have been a defining moment for generative AI (GenAI). With the public debut of tools like ChatGPT in November 2022 and rapid innovation in legal-focused AI platforms, law firms and legal departments found themselves at a crossroads: adapt or risk falling behind. This new technology promised to fundamentally alter how legal work is performed, affecting every part of litigation from mundane administrative tasks to complex strategic analysis. Has GenAI lived up to its promise?

This blog delves into this pivotal period over the last two(ish) years, exploring how the legal industry has responded to this technological inflection point. I’ll examine the pace of adoption, explore the roadblocks that have hindered widespread integration, and offer a grounded outlook on how GenAI may shape legal practice in the years to come.

The journey so far: Adoption and implementation

2022-2023: The early days

When GenAI first hit the mainstream, the legal industry’s response was a blend of enthusiasm and skepticism. There was a palpable buzz around the potential of these new technologies to streamline workflows, reduce costs, and even enhance legal analysis. The initial promise was tantalizing: Imagine AI sifting through millions of documents in seconds, drafting legal memos in minutes, or predicting litigation outcomes with unprecedented accuracy.

This period coincided with a broader industry phenomenon captured by Gartner’s “What’s New in the 2023 Gartner Hype Cycle for Emerging Technologies,” which notably placed GenAI right at the peak of inflated expectations. This meant that while there was immense excitement and interest surrounding GenAI's potential, there was also a significant risk of overpromising and under-delivering. 

For the legal industry, this translated into an environment where initial enthusiasm often outpaced practical, scalable implementation. Law firms and corporate legal teams, while intrigued, were navigating a landscape where the perceived capabilities of GenAI were sometimes divorced from its immediate, reliable utility in a high-stakes legal environment.

Despite this "hype peak," early adopters (often the more technologically forward-thinking law firms and legal departments) embarked on pilot projects, eager to experiment with AI-driven solutions for tasks like document review, preliminary legal research, and automated contract analysis. In July 2023, a survey by Wolters Kluwer of 275 legal professionals revealed that more than 80% of respondents believed that generative AI would create “transformative efficiencies.”

These pioneers saw GenAI not as a threat, but as a powerful new lever for efficiency and competitive advantage. They began testing commercially available tools, which promised AI-powered legal workflows, and even general-purpose large language models (LLMs) like ChatGPT, albeit in controlled environments. These pilot programs were crucial. They allowed teams to test capabilities, identify potential use cases, and, critically, manage the inherent risks associated with novel technology in a high-stakes environment.

Simultaneously, the legal tech vendor landscape exploded with activity. Existing players rapidly integrated GenAI capabilities into their platforms, while new startups emerged, promising specialized, legal-specific GenAI solutions. This competitive rush led to a flurry of innovation, with new features and functionalities being rolled out at a dizzying pace. Notably, DISCO had been prepared for this eventuality, delivering award winning AI-powered litigation solutions to clients since the founding of our AI lab in 2015.

However, despite this initial excitement and the rapid development of tools, implementation remained fragmented. Most law firms, especially those with more conservative leanings, largely confined their GenAI explorations to "sandbox" environments or limited pilot programs. The primary reasons for this cautious approach were legitimate: significant regulatory and judicial uncertainty surrounding AI use, particularly concerning data privacy and client confidentiality, and a deep-seated fear of misusing a technology that could have profound ethical and practical consequences. Many firms adopted a "wait and see" approach, preferring to observe how early adopters navigated the complex landscape before committing significant resources.

2024: Year 2 of GenAI

As the initial hype began to settle into a more realistic understanding of GenAI’s capabilities and limitations, Year 2 witnessed a more measured, yet still significant, expansion of AI tools within the legal landscape. The conversation shifted from "if" to "how" or “when” AI could be integrated.

This time around, in August 2024, Gartner placed GenAI lower on their hype cycle, past the peak of inflated expectations and closer to the trough of disillusionment. This shift reflected a growing awareness of the real-world challenges in deploying GenAI, including its limitations like hallucinations, the complexities of integration, and the significant ethical and regulatory hurdles. For the legal industry, this meant moving beyond the initial dazzle to confront the practical realities of making GenAI work reliably and ethically within their stringent professional requirements.

Larger legal firms, with their greater financial resources, robust IT infrastructures, and higher capacity for risk-taking, largely led the charge in adopting and scaling GenAI solutions. They began moving beyond isolated pilot projects to integrating GenAI tools into their core operational workflows. Many established internal innovation teams, dedicated to exploring, vetting, and deploying AI, or forged strategic partnerships with legal tech companies to develop bespoke solutions. This allowed them to invest in enterprise-level platforms and adapt them to their specific needs, often creating custom prompts and with their proprietary data.

Still, GenAI use wasn’t limited to larger teams with larger resources. Plaintiffs and litigation boutiques were also notably leveraging AI to gain a competitive edge against larger firms, recognizing its potential to streamline daily legal tasks and gain efficiencies.

An ABA survey at the time found that AI adoption in legal tripled over one year. In 2023, just 11% of the respondents reported using AI, citing legal research as the biggest use case. In 2024, 30% of respondents reported using AI, with  GenAI's application significantly diversifying, extending its reach into various legal sectors beyond its initial limited use cases:

• Document review: GenAI-assisted review can improve speed-to-evidence and achieve significant reductions in the time required to review documents. For instance, DISCO’s Auto Review can review documents at speeds of 32,000 documents per hour, the equivalent of a 1,920-person review team working an eight-hour day. This empowers attorneys to swiftly identify and produce essential documents, ensuring they meet their discovery obligations efficiently.
• Risk assessment and fact investigation: With Retrieval-Augmented Generation (RAG), DISCO's Cecilia Q&A empowers lawyers to swiftly get to the root of what the evidence reveals. This scalable analysis enables legal teams to rapidly uncover hidden risks and prioritize critical documents, ensuring clients receive quick risk assessments and timely guidance.
• Quality control: AI-powered QC tools can catch inconsistencies in tagging, flag potential privilege issues, and identify documents that may have been overlooked due to human error. By continuously monitoring the review in real-time, AI ensures that issues are identified and corrected early, before they escalate into costly mistakes or procedural missteps. This not only protects the integrity of the review, but also helps legal teams demonstrate defensible processes and maintain credibility with the court and opposing counsel.

The hurdles: What’s been holding legal back? 

While GenAI’s capabilities are undeniably impressive, the legal industry faces several significant hurdles that continue to impede its full and confident integration. These challenges are not merely technical. They are rooted in the culture, ethics, and structure of the legal profession itself.

Technological limitations

One of the most frequently cited and critical challenges is the inherent technological limitations of GenAI itself, particularly the issue of “hallucination.” While GenAI models can generate remarkably human-like text, they can sometimes produce factually incorrect, logically unsound, or even entirely fabricated information. In a field where precision, factual accuracy, and verifiable sources are paramount, trusting tools that might “hallucinate” is a critical flaw. Several high-profile cases have emerged where lawyers faced sanctions for submitting briefs with AI-generated fabricated citations, underscoring the urgent need for caution and judicial guidance.

Learn more about what causes AI hallucinations and how to prevent them.

Adoption resistance

A significant barrier to GenAI integration stems from within the legal profession itself. The cultural and structural resistance to change is multifaceted:

• Skepticism from seasoned lawyers: Many experienced legal professionals are deeply accustomed to traditional processes that have served them well for decades. There's a natural inclination to view new technologies, especially those as disruptive as AI, with caution.
• Uncertainty about accountability: This is a particularly thorny issue in the legal context. If an AI tool makes an error in a client-facing task or generates incorrect advice, who is ultimately responsible? Is it the lawyer, the firm, or the AI developer? The lack of clear answers here compounds the hesitancy.
• Lack of guidance from the courts: Perhaps one of the most significant external factors contributing to adoption resistance is the absence of clear, universally accepted guidance from judicial bodies on how to responsibly use GenAI in filings and court proceedings. Without such guidance, lawyers face the risk of sanctions or ethical breaches, making them hesitant to fully integrate AI into critical legal processes.

Given the high stakes and the looming uncertainty around GenAI, law firms are often in the mindset of not wanting to be either first or last to adopt, but rather, squarely in the middle.

Ethical and regulatory concerns

The legal profession is built on fundamental principles of trust, confidentiality, and strict ethical obligations. The integration of AI introduces a new layer of complexity to these existing duties:

• Data privacy and security: Can confidential client documents be safely processed using AI tools, especially those that are cloud-based and involve sending data to third-party servers? This is a paramount concern for law firms, which handle some of the most sensitive personal and corporate data
• Bias in AI: AI models are trained on vast datasets, and if these datasets reflect societal biases, the AI can perpetuate or even amplify those biases. In the context of law, this could lead to discriminatory outcomes in areas like sentencing recommendations, bail decisions, or even legal research, which would fundamentally undermine the principles of justice and fairness
• Regulatory lag: While some state bar associations, such as New York and California, have issued preliminary ethics opinions regarding AI use, emphasizing the duties of competence, confidentiality, and supervision, these are largely advisory. There remains little consensus on AI governance in legal practice at a broader, national level. Laws and regulations are struggling to evolve at the same pace as AI technology, leaving a significant void in clear guidelines for its ethical, transparent, and responsible use. This legislative and regulatory inertia leaves many legal professionals feeling exposed and uncertain about the boundaries of permissible AI application.

It is worth noting that in 2024, the ABA issued its first ethics guidance on a lawyer’s use of AI tools. This guidance provided much-needed clarity, emphasizing that lawyers' fundamental ethical obligations of competence, confidentiality, communication, and supervision apply equally to their use of AI. The guidance explicitly stated that lawyers must understand the capabilities and limitations of any AI tool they use, verify AI-generated output for accuracy, protect client confidentiality when inputting data into AI systems (especially "self-learning" ones that might use input for training), and manage client expectations regarding AI's role and potential costs. 

Notably, it stated that “[a]s GAI tools continue to develop and become more widely available, it is conceivable that lawyers will eventually have to use them to competently complete certain tasks for clients” (emphasis added). While not a definitive blueprint for every scenario, Opinion 512 served as a crucial first step in establishing a national ethical framework for AI in legal practice, influencing state bar associations and providing a foundation for future, more granular guidance.

Assessing the progress: What has worked and what hasn’t?

Despite the hurdles, there have been clear successes and demonstrable progress in the application of GenAI within the legal industry.

Successful use cases

Numerous law firms and corporate legal departments have reported tangible benefits from integrating GenAI into specific, well-defined applications:

• Time and cost savings in document review: Managing a large team of reviewers is a major cost driver in first-pass document review. However, by leveraging GenAI to quickly identify relevant documents from small samples, a smaller team can significantly streamline this traditionally labor-intensive process. This not only accelerates legal timelines but also dramatically reduces overall expenses, freeing up valuable resources for other critical aspects of a case.
• Automation of routine tasks for in-house legal teams: Corporate legal departments are increasingly using GenAI to automate routine, high-volume tasks such as legal holds, contract analysis, and discovery. This frees up in-house counsel to focus on strategic business advisory, complex litigation, and high-value legal work, thereby improving their service delivery to internal business units.
• Managing experts and reports with ease: In IP law, for example, GenAI has proven invaluable in helping cut through even the most technical jargon with accessible summaries and Q&A capabilities. Attorneys can ask questions in plain English and get answers back with a complete list of citations to documents in their own database. Attorneys can use GenAI to define obscure or unknown words or phrases in seconds, or interrogate a single document using the Q&A technology on individual files in your database.

Is the legal industry keeping up with innovation? 

When compared to other sectors like finance, healthcare, or technology, the legal industry’s pace of GenAI adoption has been notably slower. This isn't entirely surprising. Law is an inherently conservative profession, characterized by its reliance on precedent, strict ethical codes, and high-stakes outcomes. While the initial wave of AI adoption in finance, for example, saw widespread implementation in algorithmic trading and fraud detection, legal's measured approach reflects its unique regulatory and ethical environment.

However, the gap is steadily closing. What was once considered a niche area of legal tech is now a boardroom discussion. In a 2025 report by Secretariat and ACEDS, 74% of respondents expect to use AI-driven tools in their jobs within the next 12 months, indicating a significant acceleration in adoption and emphasizing AI’s critical role in the future of legal work. While large firms are leading the charge, the availability of more user-friendly and integrated AI solutions is gradually democratizing access for smaller firms and solo practitioners.

What the future holds: Trends and predictions 

Looking ahead, the next 2 to 5 years promise further evolution and deeper integration of AI within legal services, ushering in an era of unprecedented efficiency.

AI-driven legal services

Expect to see a significant expansion and sophistication of AI-driven legal services:

• Client-facing GenAI tools: The emergence of more sophisticated, user-friendly, client-facing GenAI tools is highly anticipated. These could enable more self-service for routine legal needs, such as generating simple contracts (e.g., NDAs, basic service agreements), providing personalized legal information for common inquiries, or guiding clients through basic legal processes. It's crucial to view these tools as augmenting, not replacing, legal professionals, allowing lawyers to focus on complex, high-value tasks. Careful ethical implementation is key, ensuring tools provide information and assistance, not unauthorized legal advice.
• AI copilots embedded into legal software: Rather than standalone applications, AI will increasingly be seamlessly integrated into existing legal software platforms. Imagine AI copilots embedded directly into document management systems, practice management software, and ediscovery platforms. These copilots will assist lawyers in real-time with drafting, reviewing, summarizing, and even advising, acting as intelligent, ever-present assistants.
• Cross-functional AI teams: The traditional structure of legal teams will evolve. Expect to see the formation of more cross-functional AI teams, blending the expertise of lawyers, data scientists, technologists, and legal operations specialists. These teams will be responsible for developing, customizing, and overseeing AI solutions, ensuring they meet both legal and technical requirements.

Collaboration over replacement: The human-AI synergy 

The pervasive fear of AI replacing lawyers is gradually giving way to a more nuanced and realistic understanding: the future isn’t man or machine‌ — ‌it’s man with machine.

GenAI will increasingly serve as an expert assistant, a sophisticated copilot, rather than a substitute for human legal expertise. It will free up attorneys from tedious, time-consuming, and repetitive tasks, allowing them to focus on higher-order thinking, complex problem-solving, and strategic case development. These are tasks that require creativity, empathy, ethical judgment, and nuanced understanding of human behavior, which machines cannot replicate (yet?).

This collaborative model is expected to increase lawyer productivity significantly, leading to greater efficiency,‌ lower costs for clients, and ultimately, a focus on delivering higher-quality, more sophisticated legal services. The role of the lawyer will continue to evolve, and keeping up to date with AI developments and education is paramount. As Benjamin Alaire, CEO of Blue J Legal said to Attorney at Law Magazine, “Today’s lawyers don’t need to fear being replaced – at least not by robots. They should, however, fear being outpaced (and eventually replaced) by peers who have effectively harnessed and deployed AI-powered legal tech.”

Regulatory and ethical development 

The regulatory and ethical landscape surrounding AI in legal practice will undoubtedly continue to evolve and mature. This will be a critical factor in building trust and enabling widespread adoption:

• Clearer bar association guidelines: Over the next five years, expect to see more comprehensive and prescriptive guidelines from bar associations on ethical AI use. These will likely address crucial areas such as data handling, client confidentiality with AI tools, the lawyer's duty of meaningful supervision over AI output, and responsible disclosure to clients about AI use. These guidelines will provide much-needed clarity for practitioners.
• Case law addressing AI malpractice and supervision: As AI becomes more deeply integrated into legal processes, legal challenges will inevitably emerge. We can anticipate the development of new case law addressing issues like AI-related malpractice, the standard of care for AI supervision, and liability for AI-generated errors. These judicial precedents will play a crucial role in shaping the legal profession's understanding and governance of AI.
• Industry standards for data usage, model transparency, and accountability: Beyond Bar Association guidance, we can expect the development of industry-wide standards. These will focus on crucial aspects such as the secure handling of sensitive legal data by AI systems, increasing the transparency of AI models (moving away from "black box" approaches where possible), and establishing clear lines of accountability for AI-generated outcomes. The European Union's Pioneering AI Act, for example, is already setting a global precedent for comprehensive AI regulation, which could significantly influence legal AI development and regulation worldwide. These frameworks will be instrumental in ensuring fair, transparent, and ethical usage of AI at scale.

Conclusion

The past two years have been a whirlwind for GenAI in the legal industry, characterized by both exciting promise and significant challenges. We have witnessed enthusiastic experimentation, the emergence of tangible success stories in specific use cases, and the development of sophisticated tools. Simultaneously, the profession has grappled with the inherent technological limitations of AI, deep-seated adoption resistance rooted in cultural inertia, and the complex, rapidly evolving ethical and regulatory landscape.

While the legal industry may have been slower to embrace GenAI compared to some other sectors, the trajectory is clear: GenAI is not a passing fad; it is here to stay and will fundamentally reshape legal practice. The question now is not if legal professionals will integrate AI into their workflows, but how thoughtfully, strategically, and ethically they will do it.

For legal professionals, the call to action is unequivocal: embrace AI technologies thoughtfully, understand their capabilities and limitations, and stay relentlessly informed about new developments in the field. This necessitates a commitment to continuous learning, adapting workflows, and actively participating in the ethical discourse surrounding AI. 

The future of law is not a binary choice between human and AI. It is a collaboration, human with AI. By proactively engaging with this transformative technology, the legal profession can navigate this pivotal era successfully, ensuring a future where innovation serves to enhance efficiency, reduce costs, improve access to justice, and ultimately, deliver better outcomes to clients.

Ready to explore how GenAI can transform your legal practice? Connect with us today to speak with one of our experts about tailored AI solutions for your team.

‍