California attorney general cautions tech leaders on voter deception

California Attorney General Rob Bonta cautioned executives at social media and other tech companies to work harder to protect voters from “deception, intimidation, and dissuasion,” ahead of the November election.

“Millions of Californians rely on social media and artificial intelligence services to obtain news and information about upcoming elections, and it is paramount that the platforms, products, and services offered by your companies not be misused to deceive voters about their constitutional right to vote,” Bonta wrote in a letter Wednesday to the CEOs of Alphabet, Meta, Microsoft, OpenAI, Reddit, TikTok, X and YouTube.

The letter reviewed sections of California law that prohibit interference with voting rights by misleading people about voting place and time and by using intimidation tactics.

California state law also “generally prohibits the distribution, within 60 days of an election, and with actual malice, of materially deceptive audio or visual media of a candidate appearing on the ballot in the election with the intent to injure the candidate’s reputation or to deceive a voter into voting for or against the candidate,” Bonta wrote.

The letter follows pop icon Taylor Swift’s endorsement of Kamala Harris for president on Tuesday night following the debate. Swift criticized those who had circulated AI-generated images falsely stating that she had endorsed Donald Trump.

Trump had shared a series of those images on his Truth Social platform. Separately, X owner Elon Musk recently shared an AI-generated image portraying Harris dressed as a communist dictator.

“Kamala vows to be a communist dictator on day one,” Musk, who has endorsed Trump, wrote in a post on X on Sept. 2 “Can you believe she wears that outfit!?”

Google’s Gemini, OpenAI’s Dall-E and Chat GPT, Microsoft copilot and Grok, made by Musk’s xAI, allow users to rapidly generate images and text in response to prompts or questions. In August, an updated version of xAI’s product, Grok-2, appeared to carry few limitations on creating fake images of political figures.

WATCH: The generative AI trade