September 19, 2024 - 1:00pm
California Governor Gavin Newsom has signed three bills this week targeting AI misinformation, but critics argue that the legislation will serve as a Trojan horse for censorship.
“I just signed a bill to make this illegal in the state of California,” Newsom posted on social media on Tuesday evening, referencing a satirical video purporting to be a Kamala Harris ad. “You can no longer knowingly distribute an ad or other election communications that contain materially deceptive content — including deepfakes.”
There are already federal laws in place banning the spread of election-related misinformation intended to suppress voting, which have been used in recent years to prosecute Right-wing individuals involved in a robocall scam as well as a man who flippantly claimed in 2016 that Hillary Clinton’s supporters could vote by text. The California laws go further, requiring social media companies to moderate election-related “deepfakes” and take down offending content. Critics say this will force companies to censor aggressively, taking down flagged videos regardless of whether they actually violate the law in order to avoid Government penalties.
One online parody artist has already sued California over the new laws. “Because the bills use undefined and vague statutory language, they grant California broad discretion to determine what constitutes impermissible ‘materially deceptive,’” attorneys in the case argued. “This creates a chilling effect on free speech, particularly for political commentators.”
Their client, Christopher Kohls, created the meme referenced by Gavin Newsom, a satirical Harris ad which included the phrase “I was selected as the ultimate diversity hire […] so if you criticise anything I say you’re both sexist and racist.”
Nathan Leamer, former policy advisor to the chair of the Federal Communications Commission, told UnHerd these new laws present free speech concerns by placing the burden of censorship on social media platforms.“This type of hastily written and flawed bill will only create perverse incentives for platforms to err on the side of caution, censoring and suppressing information, out of concern for being jawboned by the state government,” he told UnHerd. “Policymakers should be looking for tangible examples of harms, not hypotheticals.”
The potential for deepfakes — fake videos of real people created through AI — to impact elections has been the source of much consternation in recent years, but there has been little evidence of such videos having significant impact. Nick Clegg, President of Global Affairs at Meta, recently remarked that it was “striking how little effect AI has had at a systemic level”. Nonetheless, about half of American states have implemented or are considering laws restricting election-related AI content.
The attempt to restrict election deepfakes in California is in keeping with Democrats’ growing interest in disinformation, particularly as a cause of the party’s electoral disappointments. Hillary Clinton blamed Russian disinformation for her 2016 loss, and just this week called for those who spread such disinformation online to be prosecuted. Online disinformation has been a Democratic fixation since the 2016 election cycle, and the party has since created a Counter Disinformation Intelligence Unit.
“Where does this end?” Leamer asked UnHerd, referencing the California laws. “State policing group chats, forwarded emails from my father-in-law or Reddit comment debates?”
Join the discussion
Join like minded readers that support our journalism by becoming a paid subscriber
To join the discussion in the comments, become a paid subscriber.
Join like minded readers that support our journalism, read unlimited articles and enjoy other subscriber-only benefits.
Subscribe