OpenAI Backs California Bill Requiring Watermarking of AI Content

OpenAI, the developer behind ChatGPT, has announced its support for California Assembly Bill 3211, a piece of legislation that would require tech companies to label AI-generated content. The bill, authored by California State Assembly member Buffy Wicks, is designed to enhance transparency and reduce the potential misuse of AI-generated media, particularly in the context of political misinformation.

The Need for Transparency in AI-Generated Content

AB 3211 mandates that AI-generated content, ranging from harmless memes to potentially harmful deepfakes, be clearly labeled or "watermarked" to indicate its synthetic origin. The goal is to help the public distinguish between human-created and AI-generated content, reducing confusion and mitigating the risks of misinformation, especially during election years.

OpenAI's Chief Strategy Officer, Jason Kwon, expressed the company's strong support for the bill in a letter to Assembly member Wicks. Kwon emphasized the importance of transparency and provenance in AI-generated content, particularly as a growing number of countries, including the United States, hold elections this year. "New technology and standards can help people understand the origin of content they find online and avoid confusion between human-generated and photorealistic AI-generated content," Kwon wrote in the letter, which was reviewed by Reuters.

Legislative Context and Progress

While AB 3211 has garnered support, it has been somewhat overshadowed by another AI-related bill, Senate Bill 1047, which requires AI developers to conduct safety testing on certain models. Senate Bill 1047 has faced significant opposition from the tech industry, including from OpenAI itself. Despite this, OpenAI's backing of AB 3211 demonstrates the company's commitment to responsible AI practices, particularly in the realm of content creation.

California lawmakers introduced a total of 65 bills related to AI this legislative season, according to the state's legislative database. These bills cover a wide range of issues, from ensuring algorithmic decisions are unbiased to protecting the intellectual property of deceased individuals from AI exploitation. However, many of these bills have already failed to progress.

AB 3211, however, has seen considerable legislative success. It passed the California State Assembly by a unanimous 62-0 vote and recently cleared the Senate Appropriations Committee. The bill is now headed for a vote by the full state Senate before the end of the legislative session on August 31. If passed, the bill will advance to Governor Gavin Newsom, who must decide whether to sign or veto the legislation by September 30.

The Global Impact of AI-Generated Content

With a significant portion of the world’s population participating in elections this year, experts have raised concerns about the potential impact of AI-generated content on political processes. The prominence of AI-generated media in recent elections, such as those in Indonesia, underscores the importance of transparency measures like those proposed in AB 3211.

As AI technology continues to evolve, the need for clear guidelines and regulations around its use becomes increasingly critical. OpenAI's support for AB 3211 highlights the company's role in shaping responsible AI practices, particularly in areas where the technology intersects with public trust and democratic processes.