How will Artificial Intelligence affect the Legal Industry?

In recent years we have seen Artificial Intelligence gaining prominence across the news and in our lives. People often consider all that it can do for us – making our lives easier and completing trivial tasks – but is this all it will do? In 2021 research from Princeton revealed the legal industry as one of the most vulnerable to job loss at the hands of AI even though lawyers are often seen as requiring succinct and emotive personal skills that ‘robots’ simply cannot possess. So, is the legal industry under threat or will AI become a valuable tool for lawyers across the globe?

Where might AI take us?

AI is a very broad term often broken down into four smaller subsections: Reactive AI, Limited memory, Theory of mind and Self-aware. These four terms incapsulate where AI could go in the future. At the moment, chatbots like ChatGPT are currently considered Limited memory AI – they are able to make predictions based on past data, this is also the technology that allows for self-driving cars. Artificial intelligence does not yet understand the human mind and reasoning which is the next step for significant advancements in AI which would open the floodgates for possible AI jurors. However, current Limited memory AI is already causing discussion in the legal community. In 2023 an article from Harvard highlighted possible uses for ChatGPT starting at assisting lawyers in simple research tasks and building up to helping them with legal analysis based on imputed data. This could massively decrease the workload of barristers who have a limited time to fully analyse a case but it is not as foolproof as it seems. A New York based lawyer was sanctioned after admitting to the use of ChatGPT when creating a brief. You may think this is an overreaction but the brief contained completely fabricated citations which posed a risk to the judicial process. Alongside possible issues which can arise from the use of AI it is also necessary to consider who the machine is replacing.

Time saving vs Job saving

As people in the legal industry continue to debate the usefulness of AI there appears to be a growing number of people suggesting that AI can be used for repetitive tasks but not research. This means that many entry-level jobs in law firms and jobs for those who are newly qualified may be taken by AI. Although this automation of tasks may increase the speed and efficiency at some law firms it is necessary to question how new lawyers can learn the ins and outs of the industry if they do not start with these simple tasks. Whilst Harvard may suggest that AI can create documents and draft briefs and even top law firms like Allen & Overy have begun exploring how they can use AI, this will remove the ability for people to learn these basic but necessary skills and therefore they will become unable to ‘check’ the efficiency and accuracy of the AI in the future. Sir Geoffery Vos (one of the most senior judges in the UK) has said that using artificial intelligence could create “great opportunities for the justice system, but because it’s so new we need to make sure that judges at all levels understand [it properly]” stressing the necessity for lawyers and judges to be able to check the workings of the AI. This means that research must be conducted by the lawyers themselves because tools such as ChatGPT may use out of date or foreign law sources and unless the lawyer just redoes what the AI has done, it has not been effectively checked.

Considering Clients

A big part of being a solicitor involves meeting clients, communicating with them and keeping their information confidential. Using ChatGPT and other AI could pose a risk to this relationship in several different ways. Firstly there is the issue of confidentiality – any data inputted into a Chatbot then becomes available online and is susceptible to hacking. Leaking any information sensitive to clients may seriously damage the reputation and integrity of the legal industry. Adding onto this, Giulia Gentile (a law Fellow at The London School of Economics) wrote about the importance of The SRA Code of Conduct for Solicitors which stresses the importance of providing clients with information which does not mislead. The use of AI makes this somewhat less clear as limited memory AIs have not yet been able to create contracts with completely correct information. Gentile also suggests that clients who are exposed to what she calls ‘robo-lawyers’ and those who are exposed to faulty contracts at the hands of ChatGPT, may ultimately lose faith in the legal industry. Finally, how are lawyers supposed to make their clients feel they are being listened to if the lawyers themselves do not understand how AI actually works? Whilst the workings of AI can be complicated, surely it must be necessary to educate lawyers on how it actually operates to preserve the relationship between lawyers and their clients.

It is almost inevitable that AI will become more and more involved in our daily lives. The legal industry will have to adapt to these changes but one thing is clear: AI can not be trusted to unilaterally make decisions or craft documents for lawyers. In the future it is highly likely that small tasks will be completed using machines and that machine learning will be a key time saving tool for the legal profession. However, the legal industry must never undervalue the careful consideration of a well thought out case and they must remember that anything can make mistakes and that includes Artificial Intelligence.