Private Law Theory

ABSTRACT
Artificial intelligence is moving from the background of commercial life into the conversation itself. Today, companies make use of generative and agentic systems to answer customers’ questions, draft contract terms, negotiate prices, recommend products and, in certain cases, carry out transactions with almost no human involvement. Problems arise, however, when the machine produces something that is false. A customer might enter into a contract on the basis of an AI assistant giving an incorrect description of a product, confirming a warranty that does not exist, inventing a regulatory approval or inserting an inaccurate clause into a draft agreement. Under traditional misrepresentation law, it is assumed that a person who is legally responsible makes, adopts or otherwise is accountable for the statement in question. The emergence of AI challenges this assumption since the machine can produce the words without itself being a legal person. This paper contends that the issue should be primarily treated as one of attribution and the distribution of risk, rather than by granting electronic personhood. By drawing on common-law rules concerning misrepresentation, principles of agency, Nigerian contract and consumer protection law, as well as the Nigeria Data Protection Act 2023 and recent international developments in the area of automated contracting, it puts forward a practical framework based on control, benefit, foreseeability and the presence of reasonable safeguards. The party that uses AI as part of a contractual process should normally be held responsible for any representations that can fairly be attributed to that use, while the provider of the AI should be held responsible in cases where its own defect, misleading assurance or failure to deal with known risks results in material loss. Although the machine may be the immediate source of the words, it should not become a convenient legal ghost.

Adebayo, Faruq and Bolanta, Mariam, The Ghost in Contract: Artificial Intelligence, Misrepresentation and the Allocation of Contractual Liability (August 25, 2026).