Where's the Liability in Harmful AI Speech?

TL;DR

This paper explores the liability risks associated with harmful AI speech generated by large models, analyzing legal regimes and emphasizing the importance of technical details in liability considerations.

Contribution

It provides a legal analysis of liability regimes for harmful AI speech, highlighting the influence of technical design on liability and advocating against categorical immunity.

Findings

Liability depends heavily on algorithm design details.

Legal roadblocks exist for holding AI models and creators liable.

Courts should consider technical specifics when addressing AI liability.

Abstract

Generative AI, in particular text-based "foundation models" (large models trained on a huge variety of information including the internet), can generate speech that could be problematic under a wide range of liability regimes. Machine learning practitioners regularly "red team" models to identify and mitigate such problematic speech: from "hallucinations" falsely accusing people of serious misconduct to recipes for constructing an atomic bomb. A key question is whether these red-teamed behaviors actually present any liability risk for model creators and deployers under U.S. law, incentivizing investments in safety mechanisms. We examine three liability regimes, tying them to common examples of red-teamed model behaviors: defamation, speech integral to criminal conduct, and wrongful death. We find that any Section 230 immunity analysis or downstream liability analysis is intimately wrapped up in the technical details of algorithm design. And there are many roadblocks to truly finding models (and their associated parties) liable for generated speech. We argue that AI should not be categorically immune from liability in these scenarios and that as courts grapple with the already fine-grained complexities of platform algorithms, the technical details of generative AI loom above with thornier questions. Courts and policymakers should think carefully about what technical design incentives they create as they evaluate these issues.