After several years of delays, legislation from two Bay Area lawmakers that academics and advocates are calling a landmark artificial intelligence transparency law is set to take effect Saturday.

First passed in 2024, the law will require AI-generated content such as images and videos to contain easily accessible embedded details about its origin and history to help users identify AI-generated content. It also requires developers to provide users with a free AI detection tool.

“Three years ago, the problem hit me in the face,” said state Sen. Josh Becker, D-San Mateo, one of the bill’s co-authors, along with Assemblymember Buffy Wicks, D-Oakland, and Rick Chavez Zbur, D-Los Angeles.

Becker said learning about AI scams involving deepfakes and election misinformation galvanized him to take action.

“I realized we kind of have to come up with a way to let people know whether something was AI or not,” he said.

WATCH: Sen. Josh Becker explains AI’s impact on the workforce and the future of technology. (California Senate Democrats/YouTube)

Metadata is part of all digital content and is often used to describe details such as a book’s author or a song’s title. Under the law, AI developers must automatically record when and where a piece of content was created once it is generated. That information will remain attached to the content so viewers can trace its point of origin.

According to Becker, embedding origin details in AI content’s metadata will be harder to fake than a watermark — a visual indicator that an image or video was generated with AI — because watermarks can be added to authentic content or removed from AI-generated content.

While the law marks a step forward in transparency, Becker said the work will continue as the technology evolves. For now, he said, the legislation represents the culmination of three years of work to determine the most effective approach.

The European Union decided to delay implementation of their AI law until Sunday to align with California’s regulation coming into effect. While Europe’s regulation touches various aspects of AI regulation, California’s transparency requirements largely line up with Europe, making the laws more effective in tandem.

Becker added that enforcement will be key to ensuring the legislation is effective. Violators will face civil penalties of $5,000 per violation, enforceable through civil actions filed by the state attorney general, a city attorney or county counsel.

The next phase of the law’s rollout will begin in January, when large social media companies will be required to detect and disclose this metadata for the content they distribute while also providing users with an interface indicating whether content was generated by AI.

This story was updated at 11:53 p.m. on Wednesday, July 29.

A lifelong Bay Area resident and recent UC Berkeley graduate, Eric is inspired by journalism's ability to connect communities and hold power accountable. He plans to continue building his reporting skills before beginning graduate studies this fall.