California took a new step in the regulation of artificial intelligence. As of August 1, 2026, the first stage of Senate Bill 942, known as the California AI Transparency Act, came into effect. The law, championed by State Senator Josh Becker and signed by Governor Gavin Newsom in 2024, aims to make it easier for users to identify when content has been created or modified using artificial intelligence.
The legislation specifically aims to tackle the growth of deepfakes, digital scams, and disinformation campaigns, requiring major companies developing generative AI models to offer free tools to verify the origin of images, videos, and audio circulating on the internet.
The rule affects so-called “covered providers”—that is, companies whose generative artificial intelligence systems have more than one million monthly users or visitors and are publicly available in California. Among them are several of the main technology platforms used daily by millions of people.
The implementation of SB 942 will take place in two stages.
The first already began on August 1, 2026. As of that date, companies covered by the law must offer a free tool that allows checking whether content was generated or altered through their artificial intelligence systems.
Users will be able to upload an image, a video, or an audio file, or even paste the link to an internet post, to verify whether that material was produced with one of those models and access information about its digital provenance. Furthermore, the rule establishes that these tools cannot retain submitted files longer than necessary to perform the analysis, nor collect personal information from those who use them.
The second phase will come into force on January 1, 2027. From then on, platforms will have to incorporate visible labels that clearly indicate when content has been generated using artificial intelligence, making it easier for any user to distinguish it from authentic material.
One of the central concepts of the new legislation is so-called digital provenance data.
It consists of a set of data embedded into the content or its metadata that makes it possible to know who created it, which artificial intelligence system was used, when it was generated, and whether it was subsequently modified.
In practice, it functions as a sort of digital history aimed at facilitating the verification of content authenticity, especially in sensitive situations such as electoral processes, public health campaigns, or information that could affect user security.
The law also requires that, whenever technically feasible, AI-generated content incorporate information such as the provider’s name, the system version used, the creation date, and a unique identifier to trace its origin.
SB 942 does not only impose obligations on those developing artificial intelligence models.
It also stipulates that companies licensing their systems must ensure those tools continue to incorporate the identification mechanisms required by law.
If they discover that a third party modified the system in a way that it can no longer include mandatory tags or provenance information, they will have a maximum of 96 hours to revoke the license for use.
The legislation includes financial penalties for those who fail to comply with the new obligations.
Authorities will be able to impose civil fines of up to $5,000 for each violation, and each day the non-compliance persists will be considered an independent violation, which can significantly increase the total amount of penalties.
Actions may be initiated by the Attorney General of California, city attorneys, or legal counsels across the state’s various counties.
With this initiative, California once again positions itself as a pioneer state in the regulation of artificial intelligence, with an approach focused on offering greater transparency for users and reducing the impact of content falsified or manipulated through generative technologies.