Picture this: a world where artificial intelligence (AI) not only streamlines our everyday lives but also revolutionises the legal industry. Enter ChatGPT, OpenAI's ground-breaking innovation that is pushing the boundaries of what AI can achieve. But what exactly does ChatGPT bring to the table, and how can it reshape the future of the legal profession? In this article, I will delve into the inner workings of ChatGPT and its potential impact on the legal industry, while also addressing the limitations and challenges that come with integrating AI into the legal sphere.
What is ChatGPT?
ChatGPT is based on the advanced GPT-4 architecture and employs deep learning techniques. These techniques imitate the human brain to learn patterns and make decisions from vast amounts of data. As a result, ChatGPT can generate human-like responses to text inputs across diverse industries. By utilising a transformer neural network, ChatGPT can handle large-scale natural language processing tasks, understanding language structure, context, and semantics for accurate text generation. It is a versatile tool which can be applied across various industries. In the medical industry, for example, ChatGPT can be programmed with medical information to effectively act as a virtual physician. It can also be integrated into other systems via application programming interfaces (APIs), which allow two different systems to communicate). These APIs enable organisations to automate or streamline diverse tasks.
The impact of AI on the legal industry
Generative AI has the potential to transform the legal industry. In particular, it could have a radical effect on the way lawyers approach tasks and manage workflows in six key areas: drafting legal documents, legal research, contract review, client communication, e-discovery, and document management. AI-powered tools can effectively draft contracts, agreements, and memos, saving time and allowing lawyers to focus on more complex tasks, such as negotiating terms. AI can also streamline legal research by analysing large volumes of documents, cases, and legislation quicker than human readers. In addition, AI can automatically review contracts for potential risks and inconsistencies. Finally, AI can predict litigation outcomes based on past case data and assist during the discovery process by identifying relevant documents and information. This helps lawyers make informed decisions and reduces the time and resources required.
Although potentially enhancing efficiency, AI could negatively impact law firms operating under a billable hour model, given the reduction in the number of hours billed to clients. This might necessitate some firms to pivot to value-based billing or expand their client base. Additionally, by streamlining administrative tasks, AI could substantially reduce the role of trainees who often perform these tasks.
A prime example of AI integration in daily legal tasks is Harvey AI, an innovative artificial intelligence tool. Built on OpenAI's GPT4 technology, Harvey AI has been specifically tailored to support legal work. It was trained with general legal data, including case law and reference materials, and can be further customised using a firm's work products and templates. In February 2023, Allen & Overy announced its partnership with Harvey AI. Over 3,500 of its lawyers have already tested Harvey AI, asking more than 40,000 questions in their daily work. David Wakeling, head of Allen & Overy's Markets Innovation Group, has praised Harvey AI for its ability to work in multiple languages and across diverse practice areas, stating that “it is a game-changer that can unleash the power of generative AI to transform the legal industry”.
Although Allen & Overy is the first law firm to partner with Harvey AI, it won't be the last. Other law firms are working with Harvey to develop custom tools. In a recent development, the legal tech start-up behind Harvey raised $21 million in a Series A funding round led by Sequoia Capital. Sequoia partners Pat Grady and Charlie Curnin expressed their enthusiasm for Harvey AI on their website, describing it as the "first and best instantiation of a new breed of company: the AI super app," with the potential to transform professional services and eventually all knowledge work. With these advancements and partnerships, Harvey AI is set to have a significant impact on the legal industry, paving the way for a new era of efficiency and innovation.
Limitations of AI in the legal industry
The rapid evolution of AI and its integration into the legal industry presents three key regulatory concerns: data protection and privacy, liability, and intellectual property.
One of the most significant challenges law firms face when integrating AI into their operations is ensuring the protection of confidential client data. With data handling procedures yet to be established, many firms have prohibited the use of ChatGPT in the workplace due to concerns about potential information leaks.
Policymakers are also concerned. Italy, for instance, has temporarily banned ChatGPT while scrutinising how OpenAI collects personal information. Indeed, the effectiveness of GPT-4 relies on the consumption of vast amounts of data, which complicates the regulation of AI systems processing personal information. The European Union's General Data Protection Regulation (GDPR) generally requires a lawful basis for processing such data and mandates that it be processed solely for the purpose it was collected. Additional rules may apply to sensitive personal data. Data-hungry AI systems like ChatGPT could conflict with these requirements, which include limiting data collection, retention, and providing mechanisms for updating or deleting data upon request.
This interplay between AI systems and data protection laws poses a challenge for both law firms and regulators as they seek to balance innovation with privacy concerns. To overcome these challenges, companies like OpenAI, legal professionals, and policymakers must work collaboratively to refine AI models and ensure compliance with existing data protection laws. The way forward requires a delicate balance between allowing the legal industry to harness AI’s potential to revolutionise the profession while safeguarding privacy and maintaining trust.
The second challenge in integrating AI is ascertaining liability when the AI service inevitably fails. When AI systems eventually fail (which they certainly will), determining liability is complicated by the involvement of multiple parties in development, operation, and use. Further, AI systems and deep learning modules in particular often exhibit intricate decision-making processes that cannot be easily explained. These "black box" systems inherently do not disclose the relationship between input and output, limiting our capacity to understand, predict, and rectify issues within these systems or explain outcomes to affected individuals.
Recognising these concerns, the European Commission recently released proposals to adapt civil liability rules to the specific challenges claimants might face in seeking compensation for harm caused by AI systems. For example, to counter difficulties concerning foreseeability of harm and causation, there would be a rebuttable “presumption of causality” whereby instead of having to prove the existence of a causal link between the fault of the defendant and the output produced by the AI system, the claimant need only show that such a causal link is reasonably likely.
Harmonising these rules with GDPR will be important. GDPR mandates the provision of 'meaningful information about the logic involved' in AI systems to individuals whose data protection rights have been violated. Given the opaque nature of these systems, striking a balance between harnessing AI's potential and adhering to data protection regulations remains difficult for legislators. On the other hand, law firms may have to adapt by implementing complex contractual terms, warranties, indemnities, and limitations on liability. This view has been supported by the Center on the Legal Profession at Harvard. Andrew Perlman argues the legal profession must develop frameworks for AI systems, their developers, and users, in order to enable liability to be consistently ascertained.
The third critical aspect requiring regulation relates to the intellectual property (IP) concerns surrounding AI systems. These concerns encompass the use of copyright-protected material for training and the protection of the valuable assets generated by AI. AI systems might infringe on IP rights by utilising copyright-protected content without permission or by producing outputs that violate existing rights. For example, the rise of image generators like Stable Diffusion and Midjourney, which produce 'art' by imitating real artists' styles, has ignited a debate over whether AI-generated art is a genuine creative expression or merely a replication of existing designs. Furthermore, courts have not yet determined how protections should apply when AI systems contribute to the creative and inventive process or whether an AI system can be considered a creator or inventor.
As AI adoption grows in the legal industry, addressing these regulatory concerns becomes increasingly important. This approach will guarantee responsible and transparent use, balancing the potential benefits against the risks associated with AI integration.
How will law firms react to AI?
Law firms are likely to have varied reactions to the rise of AI, driven by the perceived benefits and potential risks. One option is for law firms to leverage first-mover advantage by immediately utilising AI. As outlined, ChatGPT has the potential to enable lawyers to work far more efficiently and to enable legal services to be provided to a larger client base at a lower price. According to research conducted by Thomson Reuters, many lawyers see AI as a complementary tool that can augment their work, producing results that neither machine nor human could achieve alone. The research reveals that 82% of lawyers believe ChatGPT and generative AI can be readily applied to legal work, while 51% think they should be. However, current adoption remains limited, with only 3% of respondents using generative AI, and 34% considering its implementation.
Law firms are likely to be cautious when integrating AI. For example, 15% of surveyed firms warn employees against unauthorised AI use and 6% have banned it outright. This caution may stem from a lack of trust in AI tools, especially regarding the handling of confidential client data. In addition to the limitations of AI previously mentioned, firms expressed concerns about AI technology causing potential job losses and economic disruption. Ethical considerations, such as the tendency of AI to perpetuate historical biases due to the overrepresentation of certain groups in their dataset, were also raised, as well as the risk of AI development being driven by market forces, which may disadvantage legal aid organisations and civil rights lawyers who serve low-income and minority groups.
In summary, law firms are likely to react to AI in various ways, ranging from complete acceptance to wholesale scepticism. While it is important to approach AI with caution, law firms must recognise the value of embracing or, at the very least, piloting AI technologies in the workplace. The competitive advantage that early adopters, like Allen & Overy, gain by incorporating Harvey AI is substantial and should not be overlooked. The potential risks of lagging behind in the rapidly evolving legal landscape far outweigh the possible drawbacks of AI adoption.
Conclusion
The integration of AI into the legal industry presents both opportunities and challenges. The potential benefits of AI, such as improved efficiency, streamlined workflows, and enhanced decision-making, can greatly assist legal professionals in their work, transforming the industry and making legal services more accessible. However, it is crucial to address the limitations and ethical considerations associated with AI, notably in data protection, privacy, liability, and intellectual property concerns. Law firm reactions to AI adoption will vary, but as AI tools like ChatGPT continue to evolve, striking the right balance between embracing innovation and mitigating risks will be essential to shaping the future of the legal industry. By carefully considering the broader implications of AI, law firms can harness its potential while navigating the challenges and uncertainties it brings.