Adobe and OpenAI Support California Bill Requiring Labels on AI Images

A 3d-rendered image of the openai logo, which consists of interlocking circles and letters, mounted on the exterior of a modern glass building under a clear blue sky.

Adobe, OpenAI, and Microsoft have backed a bill in California that requires tech companies to clearly label AI-generated content.

AB 3211 requires watermarks embedded in the metadata of not just AI-generated images but video and audio clips as well.

Despite initially opposing AB 3211, calling the bill “unworkable” and “overly burdensome”, it appears the pre-eminent tech firms have changed their mind and now support the bill which also requires social media platforms like Instagram and X to clearly mark content made by AI.

OpenAI, Adobe, and Microsoft are all part of the Coalition for Content Provenance and Authenticity which has created the C2PA digital signature system — a standard for defining image provenance which is also starting to be incorporated into cameras by manufacturers.

Reuters reports that lawmakers in California have attempted to introduce 65 bills regarding AI regulation this legislative season. Some of these were highly ambitious; one wanted all algorithmic decisions to be proven unbiased and another wanted to protect the intellectual property of the dead from exploitation by AI companies. Many of these have already failed.

But in a letter to California State Assembly member Buffy Wicks, who authored the bill, OpenAI says that it believes transparency around AI-generated content and provenance such as watermarking is important.

“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,” OpenAI Chief Strategy Officer Jason Kwon writes in the letter, which was seen by Reuters.

With a third of the world’s population heading to the polls in a year of elections all around the globe, anxiety around AI-generated content spreading misinformation is rife. Wide-held fears over a lack of media literacy among certain societal groups and unchecked synthetic propaganda spreading on social media mean there is momentum behind AI regulation.

AB 3211 will be voted on this August in the California Senate having already passed through the state Assembly by a 62-0 vote. If it passes by the end of the legislative session, it will advance to Governor Gavin Newsom to sign or veto by September 30.


Image credits: Header photo licensed via Depositphotos.

Discussion