New Jersey Advances Strong Chatbot Regulations Based on EPIC’s People-First Model

Yesterday, a New Jersey Assembly committee voted to advance A5607, a bill that would put strong privacy safeguards in place for chatbots as well as create a liability framework that will allow New Jerseyans harmed by chatbots to hold companies accountable in court.

EPIC Counsel Kara Williams testified in support of the bill to the Assembly Committee on Public Safety and Preparedness.

“This bill would limit chatbot providers’ ability to continue exploiting users’ personal data by implementing specific privacy protections over the information users give chatbots,” Williams said in her testimony. “Equally important, this bill also sets out a robust liability framework for chatbot providers.”

The bill is based on the People-First Chatbot model bill authored by EPIC, Consumer Federation of America, and Fairplay. Its privacy protections include limitations on using personal data for training purposes, to target advertisements, and to engage in extensive harmful profiling of users. Its liability system clarifies that chatbots are products for the purposes of product liability actions, states that chatbot providers are strictly liable for injuries caused by chatbots, and gives New Jerseyans the right to sue chatbot providers for harm.

The bill will next be considered by the Assembly Appropriations Committee on Thursday.

EPIC submitted written testimony in support of both A5607 and A5090 and its draft amendments. A5090 is another bill on chatbot provider liability, which was discussed by the Assembly Committee on Science, Innovation, and Technology yesterday.

EPIC thanks New Jersey lawmakers for their attention to this urgent issue and urges them to continue advancing these important pieces of legislation.