Generative AI in the Legal Realm: The Case of Varghese and China Southern Airlines

AI: The New Investment Darling in Popular Culture

Like cryptocurrency during the pandemic, a hot button issue at the moment is the growing presence of artificial intelligence (AI) in popular culture. AI is popping up everywhere, from being splashed across the front-page of the Australian Financial Review as being the new investment darling, to being spruiked by new start-ups as the basis of their business, to even the most well-established corporate actors looking to supplement their services with AI.

 

Legal Profession’s Pivot to AI and Machine Learning

Similarly, one of the more static actors in the corporate realm have started their own inroads into the mysterious world of AI, language models and machine learning . The legal profession has hardly been the most cutting-edge corporate actor in the market but with AI becoming more and more omnipresent in modern society, the notion is that the adoption of AI is necessary to provide clients with the efficiencies (time and cost) that they demand. AS we chase the benefits, which are obvious, it is not without some pitfalls that merit equal consideration.

 

Implementing Generative AI in Legal Workflows

AI use-cases have been piloted at some of the larger law firms with generative AI being used to streamline routine processes such as producing client proposals, due diligence and verification, converting large and unwieldy text into diagrams and utilising pattern-matching to unearth good pieces of precedent. The principles of generative AI to supplement legal work are sound. The ability of AI to ingest large amounts of text and interpret it provides a legal practitioner with the basis to provide their client with timely advice or help inform drafting in an efficient manner.

 

However, with any new tool comes a wealth of concerns and ethical issues to be considered, especially in the context of a risk-adverse lawyer!

 

The Ethical Considerations of AI Tools in Law

The examples of AI-use cases are examples of bespoke AI models that are built fit-for-purpose for the firms concerned. In these cases, these AI tools consider data security, accuracy, privacy and copyright in order to create a more responsible tool that is likely governed by a number of internal policies and procedures that seek to direct the end-user.

The Ethical Considerations of AI Tools in Law

The more curious cases of use (or perhaps misuse) of AI in the legal realm are well publicised and perhaps infamous among legal practitioners, highlighting the potential pitfalls that generative AI poses for lawyers.

The most prevalent generative AI is perhaps ChatGPT, which has become the one-stop-shop for many individuals that have treated it as a replacement for search engines.  However, ChatGPT has some inherent limitations and weaknesses and who better to speak on it than ChatGPT themselves! A simple prompt of “What are some issues with ChatGPT” yields a response of:

  1. Bias
  2. Misinformation and accuracy;
  3. Lack of understanding;
  4. Over-reliance;
  5. Privacy concerns;
  6. Manipulation and misuse;
  7. Context limitations; and
  8. Resource intensive.

The Curious and Cautionary Tales of AI in Legal Cases: The Notorious Incident of Varghese and China Southern Airlines

Some of these elements are of key concern for not just lawyers, but anyone working in professional services. These issues are most evident in the curious case of Varghese and China Southern Airlines Co., Ltd, a case cited by US Lawyer in Mata v. Avianca, Inc.

The issue here was that the case simply did not exist. The citation and the corresponding extract from the judgement was generated by ChatGPT following some prompting by the lawyer for some relevant cases. The lawyer’s explanation of how he came to find (or generate) the falsified case outlined a misunderstanding of the nature of ChatGPT and the lack of fact-checking in relation to the generated text.

The Evolving Role of AI and Its Future in the Legal Profession

With its tendency to hallucinate, the privacy concerns related to the input given to ChatGPT and the lack of source references, ChatGPT is not the silver bullet that many are making it out to be. However, it’s growth and improvements in sophistication present an opportunity to streamline processes, improve efficiencies and potentially, in 20 years’ time, replace roles entirely.

 

The Evolving Role of AI and Its Future in the Legal Profession: The Balance of AI Integration in Law

Ultimately, generative AI represents an epoch-setting revolution in the way we work and live. In theory, it represents the idea of a 24/7 paralegal for law firms but in reality, the need to implement safeguards to prevent breach of legal practitioner’s ethical duties and the historically slow uptake of technology in the legal market present a cooler outlook at this present point in time.

 

Whilst Madison Marcus is investigating the implementation of generative-AI in our practice, it is necessary to ensure that the privacy of our clients and the upholding of our duties as legal practitioners is of paramount importance.

PLEASE SHARE THIS

Subscribe to our newsletter