Writing for Law360, Partner Donald W. Hawthorne continues his three-part series on AI and tort liability.
The first installment of this three-part article discussed the potential application of negligence doctrine to harms caused when AI agents go rogue — taking actions that users did not authorize or intend.
That discussion identified the possibility that negligence law would not provide a recovery for all harm caused by AI agents. If the precautions taken by an AI user could be viewed as reasonable, or the harm was not foreseeable or preventable, there is a risk of a liability gap.
That gap, while present in all negligence contexts, seems particularly relevant for AI agents if, as suggested, they can be expected to misbehave in unpredictable ways, and no reasonable precaution can prevent the resulting harm.
If a negligence regime will not provide a remedy for AI agents’ harm, should strict liability be imposed and, if so, on whom?