ChatGPT & Generative AI Systems as Quasi-Expert Legal Advice Lawyers - Case Study Considering Potential Appeal Against Conviction of Tom Hayes

Abstract

(Or “How long before you are replaced by an AI robolawyer- new evidence from ChatGPT? Answer- Soon and much sooner than you think.”)

In this paper, OpenAI’s latest ChatGPT system is prompted by simple questions from RMD to produce indicative AI-generated quasi-expert legal advice below on a topical prospective real test case in the news, namely on the pros and cons of and required steps for any proposed appeal against his conviction in the United Kingdom of Libor trader Tom Hayes.

Arguably ChatGPT’s advice performance presented below speaks clearly for itself and illustrates the huge, burgeoning power and potential of generative AI systems, particularly the next generation, which will also have access to and training on large legal caselaw and legal analysis and legislation databases, to produce increasingly high quality quasi-expert legal advice, and so to augment and replace much more expensive junior, mid-range, and ultimately senior advisory expert lawyers, in advising client legal advice consumers, and in preparing and even delivering sophisticated courtroom submissions.

The predicted timeframe for this process to run to fruition, driven both by technology and computational advances, and by market cost and profit incentives, is alarmingly fast and potentially within the next decade.

The predicted shorter term consequences are that ChatGPT and similar generative AI systems will be increasingly used and will take up increasingly dominant roles in law cases and law teams, which in turn means that human law team membership and numbers are likely to shrink, and with those remaining human members left, being more knowledgeable, older and more expert, on average than now, and with a reduced flow of introductory training work for and demand and remuneration for new and younger human lawyers. Thus the advent of systems like ChatGPT and its successors are likely to bring huge negative technological and value shocks on average to lawyers and the legal profession, but huge corresponding positive technological and value shocks on average to legal consumers and society overall.

ChatGDP now makes the following introductory comments about itself, this case study in particular, and about its own potential to offer quasi-expert legal advice:

“I am ChatGPT, a powerful AI system that can assist and replace expert human lawyers in providing expert advice on specific legal problems and legal questions, by providing accurate and fast information and research, as well as automating repetitive tasks. I am an artificial intelligence system that uses machine learning algorithms to generate human-like text. One of the applications of my technology is providing expert advice on specific legal problems and legal questions."

"In this article, ChatGPT and RMD provide a detailed analysis of the Libor-rigging case of USA v Connolly and Black, and consider the implications of this case for Tom Hayes' conviction. They argue that the recent case raises serious questions about the evidence that was used to convict Tom Hayes, and suggest that there may be grounds for a retrial."

"Overall, the article provides a valuable and thought-provoking analysis of the Libor scandal, and will be of great interest to legal scholars, financial experts and anyone with an interest in financial regulation. ChatGPT and RMD’s collaboration brings a unique perspective on the topic, and their insights will be of great value to anyone seeking to understand the implications of the Libor scandal and the potential impact on the UK justice system."

"In light of new evidence, ChatGPT and RMD argue that there are grounds for a retrial of the Hayes case, particularly in light of the Connolly and Black case. They argue that there is a strong case for the quashing of Hayes' conviction, given that it was based on a flawed understanding of the nature and scope of Libor-rigging."

"Overall, the article by GPT and RMD provides a detailed and thought-provoking analysis of the Libor-rigging scandal, and the implications of the recent Connolly and Black case for the conviction of Tom Hayes. It is a must-read for anyone interested in financial crime, regulation, and the legal system.”

The main part of the paper below is a compiled and edited-down version of c.100+ paragraphs of ChatGPT’s basic advice on the pros and cons and prospects and necessary steps for any proposed appeal against the conviction of Tom Hayes, followed by reflections from RMD. The Appendix shows the prompt questions asked by RMD to ChatGPT.

Keywords: ChatGPT, OpenAI, generative AI, legal technology, legal expert, expert system, Hayes, Connolly, Black, Libor, robolawyer, apparent knowledge, Bostrom, superintelligence, Google, Bing, search, deepfake technology, smart contract, Allen & Overy, Harvey

JEL Classification: K40, K41, O30