The United States currently lacks any federal AI agent liability law, creating a legal vacuum for cases where these systems cause harm. The AI agent itself cannot be held liable as it is not a distinct legal entity, meaning liability falls on either the developer or the deployer depending on the facts and circumstances. A user who gives reckless instructions to an AI agent, such as asking it to make a hundred thousand dollars in a week without safety guidelines, may face civil and even criminal liability under the Computer Fraud and Abuse Act. This situation mirrors self-driving car accidents involving Tesla, where both the developer and the human driver can potentially be held responsible.
Source: Read the original article
