California has enacted a new law requiring artificial intelligence companies to embed invisible digital "fingerprints" into AI-generated images, videos, and audio. This measure, effective this week, aims to provide users with a verifiable way to determine the origin of digital content, distinguishing between human-created and machine-generated material.
The legislation, spearheaded by Bay Area lawmakers, mandates that AI companies embed difficult-to-remove data known as "provenance" into their creations. This data will allow content verification tools to trace the origin of files.
State Senator Josh Becker, a Democrat from Menlo Park, highlighted concerns about the misuse of AI content, stating, "While AI has some benefits, we’ve seen AI content used for election misinformation for scams and abusive deepfakes." The law addresses growing anxieties surrounding AI-generated political content, which has already appeared in California campaigns.
In response to increasingly sophisticated AI technology, U.S. Sen. Adam Schiff and Rep. Ro Khanna (D-Fremont) have previously urged Congress to regulate AI-generated campaign advertisements that could impersonate candidates or political organizations. This concern follows the use of AI-generated ads in California political campaigns, including those by Republican gubernatorial candidate Steve Hilton and former Los Angeles mayoral candidate Spencer Pratt.
Senator Schiff emphasized the bipartisan threat posed by AI-fueled political deception, noting, "AI-generated fraudulent advertising which uses the likeness or voices of candidates to misrepresent their positions, campaigns, or causes, or otherwise misrepresents a candidate’s point of view is not only wrong — it poses a serious threat to our democracy that should concern all Americans regardless of their party."
The new California law seeks to equip internet users with tools to combat digital deception. Beginning August 9, users will be able to upload images, videos, or audio files into content verification platforms, such as those developed by the Coalition for Content Provenance and Authenticity, to check if they were created by AI.
Further phases of the law are set to roll out in the coming years. By January 1, 2027, major technology companies will be required to provide users with mechanisms to identify AI-generated content within their platforms. By 2028, phones and cameras manufactured in California will be required to store provenance information at the moment an image is captured, effectively creating a digital record for each photo taken.
The development of this law took nearly three years as lawmakers worked to keep pace with AI's rapid evolution. California lawmakers collaborated with officials in the European Union, with the state law taking effect on the same day as the EU's new AI Act.
Experts suggest that while the initial phase of the law focuses on AI companies, its most significant impact could be seen when the regulations extend to social media platforms. Stephen Sharp Queener, an academic researcher at Starling Lab, indicated that the platform requirements could fundamentally alter how individuals consume information online, potentially reshaping the information economy.