Will ChatGPT Disrupt the Legal Industry?

Every lawyer aspires to provide the best service to their clients and every client looks at the value arbitrage for the fees paid. With the rise of Artificial Intelligence (AI) and automation, there is an opportunity for lawyers to automate routine and monotonous tasks and this will be of benefit to clients.

Artificial Intelligence (AI) is certainly a direction forward and through this article, we will discuss how OpenAI’s new offering ChatGPT might disrupt the legal industry, but let’s first understand the technology.

Chatbots are Artificial Intelligence (AI) systems that use natural language processing to understand and to respond to human communication. In general, there are two types of chatbots: retrieval and generative, of which ChatGPT is the latter. While generative chatbots employ an underlying deep learning model to construct the output, retrieval chatbots identify user input patterns and offer a prepared response.

A strong language model-based chatbot like ChatGPT could enable lawyers to automate necessary but time-consuming tasks. Lawyers could spend their valuable time finishing and polishing and then move on to other important tasks instead of doing legal drafting by using an advanced language model to generate draft contracts or briefs swiftly and efficiently with minimal scrutiny. Similarly, such a model may concentrate its language analysis to glean vital information about precedents and practises from voluminous textual material.

Unknown Opportunities

Omer Tene, Technology group and Data, Privacy & Cybersecurity practice Partner with Goodwin Procter, recently posted on LinkedIn his experience with ChatGPT to draft a privacy policy. Omer’s post stated, “It wrote a really nice one. In two seconds (non-billable). Any idea what lawyers will do next?”

Chatbots and Generative Pre-trained Transformer (GPT) models like ChatGPT have the potential to greatly impact the legal industry. A few ways in which they could do so:

1. Automating Legal Research: GPT models can be trained on legal text and used to generate summaries of legal documents, helping lawyers quickly find relevant information and make more informed decisions.
2. Generating Legal Documents: GPT models can be trained on existing legal documents and used to generate new ones that are similar in style and content. This could be a time-saving tool for lawyers and legal teams, allowing them to quickly generate contracts and other legal documents.
3. Legal Chatbots: Chatbots can be used to provide information and answer questions for clients, reducing the need for lawyers to spend time answering routine questions. This allows lawyers to focus on more complex tasks and improves the efficiency of legal services.
4. Legal AI: AI-based systems have already begun to be used in legal discovery and document review, helping to identify relevant information quickly, and this trend is set to continue.

All these points however would require training GPT models on accurate legal texts and data, as they are Generative by nature, they will generate by taking a contextual examples as input, GPT model will be as good as the dataset it is trained on. Additionally, the final outcome would be determined by the specific task you want the model to accomplish, the quality of the data, and the skill of the developer, as well as the ethical considerations which must be considered when developing AI systems.

Even the least technically adept person will find ChatGPT simple to use. Asking a query at the text input prompt is all it takes to communicate with the chatbot.

It's worth noting that while AI has ample potential to improve the legal industry, it also raises a number of ethical and legal questions, like issues of accountability and fairness. It is important for legal professionals to be aware of these issues and take steps to address them as they begin to incorporate AI into their work.

Deployment is a concern

Because of issues with cost, security, and trust, deployment of AI applications has historically been difficult in the legal sector. With ChatGPT, it most likely won't change. While generative models offer many opportunities, they also raise a number of ethical and security issues, most of which are connected to ChatGPT's intricate structure.

For starters, if they utilise chatbots in their practice then lawyers have a duty of being armed with technological knowledge and should be aware of both the chatbot's advantages and disadvantages and attorneys who choose to employ it must be able to communicate to their clients of the benefits and drawbacks of doing so. The possible risk of data bias associated with any AI technology is another issue. If client data is kept by third parties, confidentiality concerns also come into play.

Also, most legal enterprises and departments do not have the necessary hardware to manage data at a volume that would produce outcomes similar to those of GPT-like models inside their secure firewall. Therefore, using ChatGPT or a comparable tool must pay close attention to inputs coming from outside their network. The amount of data needed to deploy these AI models rapidly increases the difficulty of security and will most likely be rather overwhelming for most organisations.

Lastly, being a machine learning system, ChatGPT cannot have the same understanding and judgement when it comes to interpreting legal principles and precedents akin to a human lawyer. This could cause issues in circumstances wherein a much more in-depth legal study is necessary. 

AI takes over the role of Lawyers

In his article for Reuters analysing ChatGPT, Andrew Perlman, Dean at the Suffolk University Law School mentioned a quote from the bot creations. The quote on the OpenAI portal states, “ChatGPT shouldn’t be relied upon for advice, and that it sometimes writes plausible-sounding but incorrect or nonsensical answers.” 

If a lawyer did that, there could be malpractice consequences - but if the bot steers you wrong, there’s not much that can be done about it, other than it being your very own personal loss or damage. To put it in simple words, there’s no accountability for the accuracy of the information provided which simply will not work in the legal industry.

So, for the foreseeable future, chances of ChatGPT or any other such AI program replacing lawyers seem slim but we cannot outright strike off the possibility of this ever happening considering the rapid pace at which AI tech is advancing.

Perlman agrees that’s a concern. But he doesn’t see it as an either/or situation. Lawyers could use the technology to enhance their work, he said, and produce “something better than machine or human could do alone.”

To conclude, ChatGPT has the potential to be a game-changing add-on to the legal industry but its gambit will only be limited to automating repetitive and mundane tasks which no doubt will still be a big help and would result in a lot of man-hours saved however for anything more complex such as legal advisory or complex transactional drafting or deal structuring, it still seems a bit far-fetched and too much for any AI technology to handle.

Author: Varun Bhatia, Co-Founder of 3NServe.

Also read top viewed Ai Legal article: The Role of AI in Legal Research.