Attribution of pre-contractual information and knowledge : the ELI Model Rules on Digital Assistants for Consumer Contracts and beyond

The potential of using artificial intelligence (AI) systems to automate some or all of the stages of a typical contracting lifecycle has been much debated, both in scholarly literature and in the development of model rules and principles. In these discussions, much attention has been paid to how one should categorise an AI system used for contract automation, and to what extent the actions of an AI system should be attributed to the person who has used it to automated their contracting activities. A related, but distinct, question concerns whether, and to what extent, any information or “knowledge” received by an AI system should be attributed to the person using it for its contracting activities. This might be significant insofar as the knowledge of a contracting person might be relevant to the application of some doctrines of contract law, as well as because the law frequently imposes an obligation on one (prospective) contracting party to disclose information, or to give information, to the other party. The purpose of this contribution is to examine how the European Law Institute tackled this question in its Guiding Principles and Model Rules on Digital Assistants for Consumer Contracts (DACC). The DACC are the first set of model rules that specifically focus on contract automation in a consumer context. As legal information duties are particularly prevalent in consumer law, focusing on the DACC is instructive, although there will also be a comparison with the other two key texts, UNCITRAL’s Model Law on Automated Contracts (MLAC) and the Principles for AI in Contracting (PAIC).