Negative Obligations in Policymaking: Contractualist Approaches to Generative AI

Abstract

The future of generative AI (GenAI) presents various ethical challenges for lawmakers, as effectively legislating a rapidly evolving technology can be difficult when so much of its implications are still unknown. By introducing a conceptual analysis of what positive and negative obligations are appropriate for present and future people, the article will argue that we owe NO to both present and future peoples; that is, we bear the duty of ensuring human survival and taking precautionary action on behalf of all generations. However, while we owe PO to living people, we do not owe those same obligations to future generations due to the future’s temporal remoteness and moral vagueness. Policymakers, therefore, should also operate based on these principles of positive and negative obligation, an approach which is justified by contractarianism. Towards the end, it will engage with a case study on the EU AI Act and show how it links with a suggested policy making process.