ABSTRACT
Contracts made by algorithms with the capacity for artificial intelligence pose a problem of correspondence with contract law, particularly in regard to the doctrine of mistake. Artificial intelligence permits an unprecedented level of autonomous contract formation without human awareness, meaning that the traditional approach to setting aside contracts for unilateral mistake may well be unworkable. This article proposes an update to the doctrine of mistake for the purposes of artificial intelligence. It is suggested that contracts formed with artificial intelligence should be set aside for mistake when two conditions are met: 1) the transaction produces an unconscientious advantage or detriment for one of the parties; 2) the claimant machine owner does not bear greater responsibility for the mistake or its reasonable avoidance. The second step would consider the efficient allocation of risk, and whether a mistake was endogenous or exogenous to the computer program. While sound policy would dictate that all parties contracting through artificial intelligence exercise appropriate diligence in the design of their programs, it is proposed that endogenous mistakes should generally be treated more strictly than mistakes that are exogenous.
€
Michael Ilg, Rage against the Machine: Artificial Intelligence and Mistaken Contracts 70 Canadian Business Law Journal 59 (2025).
Leave a Reply