James Toomey (University of Iowa – College of Law) has posted Zombies, AI, and the “Objective” Theory of Contracts (Harvard Journal of Law & Technology (forthcoming 2026)) on SSRN. Here is the abstract:
“Agentic” “AI” that can pass the “Turing Test” might behave in ways externally indistinguishable from persons but without our subjectivity or mental states. Under the so-called “objective” theory of contracts, taught as black letter in the common law world for a century, this is purportedly enough for large language models to enter into self-binding contracts. In contrast, this Chapter argues that this possibility offers a reductio in favor the alternative view that takes the “objective” theory as the law’s evidentiary commitment rather than a normative or metaphysical one. For starters, it is not clear that language “agreed upon” by entities without mental states means anything, at least not anything anyone cares about. More generally, there is no normative reason for the law to attempt to enforce the output of probabilistic language models entirely disconnected from subjective intent; the only possible justifications of contract are, in the end, “subjectivist.”
For a different perspective, see Legal Personhood for Artificial Intelligences which argues for a functionalist approach to legal personhood and by implication capacity to enter into contracts. On the question whether AIs can create meaning, see Artificial Meaning, which demonstrates the possibility that an AI with the requisite functional capacities could produce “meaning.” The author does not consider the arguments presented in either piece.