Catastrophic Liability: Managing Systemic Risks in Frontier AI Development

TL;DR

This paper discusses the systemic risks posed by advanced AI development, highlighting transparency issues and proposing a comprehensive safety and liability framework inspired by other high-stakes industries.

Contribution

It introduces a novel safety and liability management approach for frontier AI, adapting frameworks from nuclear, aviation, cybersecurity, and healthcare sectors.

Findings

Highlights transparency gaps in current AI safety practices

Proposes a comprehensive safety documentation framework

Suggests liability structures for AI-related harms

Abstract

As artificial intelligence systems grow more capable and autonomous, frontier AI development poses potential systemic risks that could affect society at a massive scale. Current practices at many AI labs developing these systems lack sufficient transparency around safety measures, testing procedures, and governance structures. This opacity makes it challenging to verify safety claims or establish appropriate liability when harm occurs. Drawing on liability frameworks from nuclear energy, aviation software, cybersecurity, and healthcare, we propose a comprehensive approach to safety documentation and accountability in frontier AI development.