Guest post by Simon Gibson, solicitor and chief executive of the Spirant Group
The legal landscape is perpetually changing, with artificial intelligence (AI) being the most recent frontier. Law firms are being swayed by the promises of AI to automate routine work, deliver faster results and potentially provide greater value to clients.
However, a recent incident highlights that while AI advancements are undoubtedly exciting, caution must be exercised, underlining the importance of ongoing education on this new technology’s potential and limitations.
The Avianca case
In a lawsuit against Colombian airline Avianca in the New York District Court , the plaintiff’s lawyers were caught submitting fabricated case citations supposedly generated by ChatGPT (the AI chatbot released by Open AI in November 2022), according to the New York Times.
The defence team identified these bogus cases. The lawyer, Steven A Schwartz, was fooled by the AI’s claims of case authenticity and apologetically conceded his errors.
Mr Schwartz had, before submitting the false case citations, taken the eccentric step of challenging ChatGPT as to the veracity of the case names it produced. The AI remained adamant that the cases were genuine, referring Mr Schwartz to Westlaw and LexisNexis.
It appears that Mr Schwartz did not investigate further and accepted Chat GPT’s position without demure and submitted the case citations to court and his opponent shortly thereafter.
The court will shortly call a sanctions hearing at which Mr Schwartz will have some explaining to do.
What AI is and what it isn’t
It is crucial to understand that AI is a tool designed to help and support us, not to replace human expertise or judgment. AI, especially in the field of law, is still in its formative stage. While impressive in many respects, it isn’t fool proof and is clearly susceptible to errors.
ChatGPT was trained on a diverse range of internet text. However, it does not have access to an up-to-date, exhaustive, or accurate legal database. It is not designed to be relied upon for legal advice, let alone case citations.
This incident has underscored the significance of understanding the functionality, reliability and limitations of AI before integrating it into professional practices.
The law is intricate, complex and forever evolving, hence the reliance on the human ability to interpret, strategise and present it accurately. AI, as it stands today, lacks the capacity to comprehend context fully, interpret nuances, or exercise judgement. It cannot replace years of legal expertise, understanding and the ability to engage in sophisticated legal thinking.
In addition, as Mr Schwartz has discovered, there are times when the AI is just plain wrong.
Education, not application
Therefore, the message is clear: the next couple of years should be about the legal sector continuing to learn about AI and its benefits, rather than rushing to purchase or implement it prematurely. AI is not a panacea for all legal challenges; instead, it should be seen as an evolving tool designed to augment, not replace, human capabilities.
The legal fraternity must take the time to understand the functions of AI, its potential and its limitations. There is a wealth of high-quality books, courses and other media to assist us in familiarising ourselves with this emerging technology.
We must also look to collaborate with technology experts, to understand how AI can be safely, ethically and effectively incorporated into our work.
Investing time in understanding AI now will serve the legal community well in the future. This area of technology will rapidly advance in the years ahead and we will need to be prepared to evolve our practices in synergy.
However, at least in legal work, it is likely that the AI will always need to be underpinned by a knowledgeable, human touch.
In conclusion, while the allure of AI’s promise in law is enticing, this Avianca case should serve as a wake-up call to the profession.
We are in a transformative era of legal practice, and as with any transformation, understanding and adaptation is key. AI can be a useful addition to the legal toolbox, but it should be handled with care, caution and the backing of comprehensive understanding.