A Santa Clara mayoral candidate has pulled a batch of AI-generated images from his campaign website after a political consultant accused him of violating a state law that requires disclosure of artificial intelligence in political ads. Gary Ferraris took down the images on Friday, just after being notified that a complaint had been lodged against him with the California Fair Political Practices Commission.
Diana Torres, a freelance political consultant, filed the complaint with the FPPC on Sept. 24, alleging that Ferraris' campaign website used undisclosed AI-generated images, according to San José Spotlight. The images depicted AI-generated campaign volunteers and supporters, including a person canvassing at a campaign table, a construction worker holding a Ferraris campaign sign, and Ferraris posed with a supporter wearing an “Engineers for Gary Ferraris” shirt, per the outlet's report.
One of the images also showed a building labeled “City of Santa Clara” that, according to the complaint, does not resemble the city's actual City Hall. Torres said Ferraris' website contained fake workers, fake volunteers and fake supporters, and argued that voters should be able to trust that campaign images of people and events are real.
Ferraris Calls It an Oversight, Not an Attempt to Deceive
Ferraris, a former Santa Clara maintenance worker and labor representative, is one of six candidates running for mayor in the Nov. 3 election, alongside Karen Hardy, Kevin Park, David Kertes, Kathy Watanabe and Mike Wiltberger. He characterized Torres' complaint as a political attack, but also accepted responsibility for the oversight, saying an honest mistake should not become a political distraction and that his campaign is focused on issues important to residents, the same account states.
The images were removed from the campaign site the same day the FPPC notified Ferraris of the complaint, according to the report. As of publication, the FPPC has not publicly confirmed whether it has opened an investigation into the matter. Shery Yang said the commission generally confirms receipt of complaints two days after respondents are notified, per the outlet's reporting.
The Law Behind the Complaint
The complaint rests on Assembly Bill 2355, which took effect Jan. 1, 2025, and requires any political advertisement containing an image, audio or video generated or substantially altered using artificial intelligence to include a disclosure. The bill text would require a committee that creates, originally publishes or originally distributes a qualifying political advertisement to include a specified disclosure that the material was generated or substantially altered using artificial intelligence. Separately, the FPPC notes that candidates and political committees must generally put disclosures on campaign ads identifying the committee that paid for or authorized the communication.
A $645,000 Wrinkle From the 49ers
Torres also pointed to an independent expenditure sponsored by Forty Niners Football Company LLC that reported $645,000 in spending supporting Ferraris and opposing Watanabe and Kertes, saying Ferraris has that sum in special interest spending from the 49ers behind him. That money does not go directly to Ferraris' campaign, though it can be used for ads, mailers and other forms of campaign communications, per the same report.
Torres, who previously worked on external engagement strategies for the San Francisco 49ers and has supported independent expenditures herself, said her past work for the team does not change her concerns about Ferraris' AI-generated imagery.
Part of a Wider Pattern This Election Cycle
The Santa Clara dispute lands amid a broader surge in AI-generated political content nationwide. The Wesleyan Media Project identified at least 164 political ads generated or enhanced with AI during the 2026 election cycle, and 69% of them did not disclose that use, according to the Business Journal. Roughly 80% of those ads were sponsored by Republican candidates or pro-Republican groups, the outlet reported, and even in states with disclosure laws on the books, only 29% of AI-related ads actually carried a disclosure.
Enforcement of these new rules remains murky. Experts cited by the Business Journal said it is not yet clear how enforcement works or what exactly falls within the purview of these disclosure laws. New York offers a parallel case: its state Democratic Party filed an August complaint alleging Republican gubernatorial candidate Bruce Blakeman violated a state law requiring disclosure of AI-generated images in political communications.
National spending on AI in campaigns has also climbed. Federal candidates and campaigns have reported at least $17 million in disclosed spending on AI technology vendors across 523 campaigns since 2020, according to the Guardian. The publication also noted that a September 2025 Pew survey found more than 70% of Americans would think less of a candidate who used AI to help write a speech, a data point that underscores why disclosure fights like the one in Santa Clara carry real political weight.
