OpenAI Backs California’s AI Transparency Bill

OpenAI has come out in support of a new bill in California that would require companies to better identify content created or augmented with generative AI.

Bill AB3211, drafted by Buffy Wicks, is one of 65 bills drafted in California this legislative season that target regulating AI.

Bill AB3211 aims to ensure that companies “apply provenance data to synthetic content produced or significantly modified by a generative AI system that the provider makes available.” In other words, content, whether images, video, or audio, created or significantly modified by AI needs to have built-in data, such as a watermark, to allow others to identify its source.

In addition to creating this metadata, companies will have to use techniques that make it difficult to remove the identifying information, and the bill prohibits the creation and distribution of software that can remove or change the data.

To provide even more transparency, certain platforms, such as Instagram or X, will have to translate the metadata into labels that are easy for users to read and understand.

This bill is still working its way through the various approval steps. However, on its first vote in the assembly, it passed 62-0. It is now waiting for its third reading in the State Senate, with a target date of July 1, 2026, for the bill to come into effect.

When the bill was first introduced in April, many tech companies, including a trade group representing Adobe and Microsoft, opposed it, calling it “overly burdensome” and “unworkable.”

After recent changes, OpenAI, Adobe, and Microsoft are now supporting the bill.

With nearly one-third of the world having major elections this year, regulators in many countries are looking to AI, how it can impact the democratic process, and what regulations can do to protect elections.