“Ordinary Meaning” in an AI World

Last month, Judge Kevin Newsome of the Eleventh Circuit authored a concurring opinion in which he discussed his use of artificial intelligence engines to considering the “ordinary meaning” of the word “landscaping” as used in an insurance policy. While the opinion itself is an interesting read (and a refreshing break from usually turgid legalese), the underlying concept explored by Judge Newsome has potentially significant ramifications for contract drafting and interpretation. Specifically, the judge wrote that because AI engines “train” or “learn” by reviewing billions of words of “ordinary” or “common” language available on the internet, an AI’s “understanding” of any particular word is very likely the consensus understanding of the ordinary meaning of that word. The judge also acknowledged the obvious shortcomings and pitfalls of simply relying on AI to define words – and I would add that using an AI’s interpretation of more loaded terms or phrases based on internet inputs seems particularly fraught – but suggested that they might be useful as a tool, among others, to interpreting contracts.

So, while it is unclear whether or when Judge Newsome’s approach may be adopted by other courts, careful practitioners should consider running key “ordinary” terms in their contracts through an AI engine or two just to make sure that the internet’s understanding of your “ordinary” term is not significantly different from what is intended.