Lawmakers call for crackdown on AI deepfakes after Grok backlash

A group of Democratic lawmakers are pushing the Federal Election Commission (FEC) to increase regulation on artificial intelligence deepfakes following the release of the social media platform X’s AI chatbot Grok.

In a letter to the FEC on Monday, Rep. Shontel Brown (D-Ohio) and a handful of other House members asked the regulatory agency to clarify whether AI-generated deepfakes of election candidates are classified as “fraudulent misrepresentation.”

In doing so, the lawmakers threw their support behind a July 2023 petition by Public Citizen, a progressive consumer rights watchdog group. The group asked the FEC to propose rules on deceptive AI in campaigns amid rising concerns about the dangers deepfakes pose during the election season.

The advanced technology can generate images, audio, and video and digitally alter likenesses and voices.

“The proliferation of deep-fake AI technology has the potential to severely misinform voters, causing confusion and disseminating dangerous falsehoods,” the letter wrote. “It is critical for our democracy that this be promptly addressed, noting the degree to which Grok-2 has already been used to distribute fake content regarding the 2024 presidential election.”

The lawmakers said this comes after X’s launch of its AI image generator Grok-2 image generator, released earlier this month with seemingly few guardrails.

The House members discussed how fake images of presidential candidates Vice President Harris and former President Trump have circulated on X and other social media platforms in the weeks following the release. They pointed specifically to Trump’s sharing of several fake images of pop star Taylor Swift and her fans supporting his campaign earlier this month, though it was not immediately clear what platform was used to generate the images.

Other signatories on the letter include Democratic Reps. Eleanor Holmes Norton (D.C.), Greg Landsman (Ohio), Summer Lee (Pa.) and Seth Magaziner (R.I.)

Sean Cooksey, the Republican chair of the FEC, pushed back against the letter’s demands, writing, “Our First Amendment rights don’t disappear when we use an AI image generator to express ourselves.”

“While the FEC will continue to enforce the campaign-finance laws on the books to promote transparency and accountability, I oppose any effort to suppress political speech by regulating a technology that few agencies even understand,” he wrote in a statement to The Hill. “Our elections should be decided at the ballot box after a free exchange of arguments and ideas.”

Cooksey has been a vocal opponent of efforts to have the FEC regulate AI and wrote in an op-ed earlier this month the agency “has no business” in passing rules to do so for political campaigns.

He proposed halting any rulemaking on AI with no further action. The proposal was supposed to be voted on during an August 15 meeting, but it was canceled. The commission is again slated to take up the proposal during Thursday’s meeting.

The Hill reached out to X for further comment.

Grok is only available to subscribers of the premium versions of X, though some have expressed concerns the content is being shared across multiple social media platforms.

Earlier Tuesday, Grok was updated after five secretaries of state warned it was spreading false information about the election. After President Biden announced his withdrawal from the 2024 race last month and endorsed Harris in late July, the AI chatbot suggested the ballot deadlines already passed in a handful of states.

Now, when people use election-related terms in their inquiries to Grok, it now directs them to Vote.gov, a nonpartisan government website, Minnesota Secretary of State Steve Simon said in a release.

Since Elon Musk purchased then-Twitter in 2022, watchdog groups and some lawmakers have mentioned concerns over a potential increase in hate speech and misinformation on the platform given its loosened content moderation policies.