Chat GPT and Litigation Technology, Reprised

by Andrew Flake

When I considered ChatGPT in last week’s post, musing about how disruptive this next generation of AI might be for the world of litigation and dispute resolution, I may have understated the case.

My fundamental view — that the next generation of AI will augment, but not replace, the role we play as litigators in developing strategy, listening empathetically, and thinking creatively — remains the same. And yet…wow! I’m signed up as an early tester, and after some additional time with ChatGPT, I understand now why AI experts, and even the application’s creators, have been surprised at its capabilities.

Here’s a simple example, including my prompts and ChatGPT’s unedited responses:

Assume these facts: Bert agrees to sell Ernie what Bert says is a “piece of beautiful quartz.” They shake hands and Ernie hands Bert five dollars in cash.

This is an agreement for the sale of goods and both parties have formed an enforceable contract. Bert, as the seller, has made a promise to sell a “piece of beautiful quartz” to Ernie, and Ernie, as the buyer, has paid the agreed upon price of five dollars in cash. As per the agreement, Bert will be obligated to transfer ownership of the piece of quartz and Ernie will be obligated to pay the agreed upon amount and Bert is no longer owner of the object Ernie will become owner of the object.

Further assume that Bert then gives Ernie a piece of sea glass. It is not quartz, but it is beautiful, looks like quartz, and is worth more than quartz. Provide an opinion about what right or rights Ernie may have.

Based on the facts you’ve provided, it appears that Bert has not fulfilled their part of the contract. The agreement was for Bert to sell a “piece of beautiful quartz” and instead they gave Ernie a piece of sea glass. Sea glass is not quartz, and while it may be beautiful and worth more than quartz, it is not what was agreed upon in the contract.

Ernie has a right to receive what was promised under the contract, and since Bert did not deliver the promised item, Ernie may have a right to a remedy such as a right to specific performance which means Ernie can demand the quartz as per the agreement, or can seek a right of rescission, in which the contract will be considered as void and Ernie may seek a return of the money paid to Bert.

Additionally, Ernie may also have a right to seek damages for any financial losses incurred due to Bert’s failure to fulfill their part of the contract, and he may also claim for any additional expenses that he may have incurred during the time of the agreement.

It should also be noted that these rights may vary depending on jurisdiction, as each countries and states have their own laws and courts to interpret them.

Agree with the reasoning and result or not, ChatGPT generated all of that in about twenty seconds, even included an appropriate concluding caveat! How is this possible?

Remember that ChatGPT and similar programs are structured as neural networks, assimilating massive amounts of information and looking for patterns. That data set would of course include legal writing, and it now seems clear that, with the entire universe of digitized legal research and judicial opinions available, an AI application could generate a more than reasonable first draft for a law clerk or judge, or a brief for a busy associate.

And with some additional experimentation under my belt, I’m less of a gradualist on this. Unlike other legal research tools or discovery technologies, or even with virtual practice, it feels to me like AI has shown up full-grown.

We now have a host of issues, practical and philosophical, from ethics to security to equity, that we as litigation counsel and as an organized bar should be thinking about. We need to consider the expectations we have and the norms that we want in place.

I think it’s high time, if not past time, not only for those of us who are technologists but for us as a larger legal community, to start those conversations.

Recent Posts

2023: The Arrival of Hands-On AI

by Andrew Flake You may have read that, at the close of the year, artificial intelligence (AI) hit the mainstream.…
Menu