OpenAI, developers of ChatGPT, has supported a California bill to make technology companies label AI-generated content. This follows the state’s concern about how deepfakes might affect public trust and confidence during election periods. The artificial intelligence transparency bill, AB 3211, mandates that contents created using AI be “watermarked.”
The bill, which is already being driven by Buffy Wicks, a member of the California State Assembly, has gained notable traction in the state legislature. It was passed through the Senate Appropriations Committee and received unanimous support from the state Assembly. As it now heads for a full Senate vote, AB 3211 remains one of the few pieces of legislation relating to AI still under consideration this year in California.
California lawmakers drive new AI accountability measures
The move to push AB 3211 is part of a wider attempt by the legislature to regulate AI in California. This year, legislators put forward 65 bills touching on different angles of artificial intelligence. However, many of them have died out, making AB 3211 still an important bill within this ongoing debate about the role AI should play in society.
OpenAI has thrown its weight behind it by supporting the legislation that underscores the significance of transparency concerning AI-generated content. OpenAI’s Chief Strategy Officer – Jason Kwon wrote to Assembly member Wicks in support of the bill and emphasized that new techs and standards must be developed to allow people better understand where the content they encounter comes from. He admitted concerns being raised during electioneering years due to AI-made materials appearing.
Reuters reviewed Kwon’s letter, which said that new technology and standards could help individuals identify the source of online content and prevent confusion between human-generated and AI-generated content.
California lawmakers tackle AI regulation challenges
While AB 3211 has made some headway, another AI bill in California, SB 1047, has been met with significant resistance from technology industry players such as OpenAI. According to many stakeholders in the AI sector, this legislation would be too burdensome. On a broader level, this is one flashpoint in the larger discussion around how to regulate AI and shows how lawmakers are grappling with striking a balance between innovation and public safety.
Unlike SB 1047 which has attracted a lot of controversy, AB 3211 is not controversial either since it focuses more on accountability and transparency matters. The bills aim at addressing one of the most pressing issues related to AI which includes synthetic media’s ability to deceive people by requiring their contents be labelled.
This will see it go through Governor Gavin Newsom’s desk if AB 3211 passes the Senate by August 31 at the end of the legislative session. On or before September 30, Newsome will have to either sign it into law or veto it. This legislation would make California one of the first states to require such transparency measures concerning content created by AI setting a precedent for others followed perhaps by federal regulation.
