Elon Musk, who has pledged his support to Republican former President Donald Trump, does not seem to face any limit to how much he can use social media to tilt the political field. He is facing accusations in the U.K. that he’s fanning the flames of anti-immigrant riots with incendiary comments on X, the platform he owns. Last Friday, he shared a doctored version of Vice President Kamala Harris’ campaign ad where her voice had been modified with AI to call herself a “diversity hire”— defending the X post as “parody” even as it sparked concerns about misleading uses of the technology ahead of the election. On Monday, five secretaries of state — who act as the chief election officers for Minnesota, Michigan, New Mexico, Pennsylvania, and Washington — pushed Elon Musk in an open letter to course-correct after Grok, X’s AI search assistant, produced false information for more than a week. Grok erroneously told users that the ballot deadlines for several states, including theirs, had already passed. “X has the responsibility to ensure all voters using your platform have access to guidance that reflects true and accurate information about their constitutional right to vote,” the secretaries wrote. They urged Musk to have Grok direct users to a nonpartisan website (CanIVote.org) operated by the National Association of Secretaries of State if they ask it election-related questions — something that ChatGPT already does. Experts say they fear what the letter reveals: with Congress stalling and a patchwork of state law straining to keep pace with AI advancements, officials have little recourse to combat misinformation in the final months before the election — except to make a last-ditch request of the platforms themselves. “You have secretaries of state begging Elon Musk,” said Sarah T. Roberts, an information studies professor at the University of California, Los Angeles and former staff researcher at what was Twitter. “What that suggests is a total failure of any legislative apparatus to appropriately go after this.” Minnesota’s Democratic Secretary of State Steve Simon, who led the letter, said it was meant to be narrowly focused on preventing election misinformation and not directed at any of Musk’s other recent controversies. He told DFD, “I certainly would be undertaking this effort if any AI chatbot of significance owned by anyone in any company anywhere were doing the same thing.” “It's important to speak up loudly now because a similar mistake in the future, over the next 92 days, might be a higher stakes situation,” Simon said in an interview. “Today, it's about ballot access, but if left unchecked and unpoliced, what if next month it's about voter registration rules?” Simon said he recruited secretaries to the letter after X shrugged off the National Association of Secretaries of State when it reported Grok’s problem. He said the company confirmed that it had received the secretaries’ letter but has not indicated whether it would comply. X gave DFD an automated reply to questions about Musk and the letter: “Busy now, please check back later.” All year, election officials have warned about the threat of AI-generated misinformation. But Grok, Simon said, represents a new kind of challenge. “What makes this particularly significant is that this isn't just an example of some other person, a random user or group of users saying something that's inaccurate on the platform. That happens regularly,” he said. “We're asking them to correct their own mistake, their chat bot, their AI-driven tool.” Roberts said that beyond cutting down on the worst abuses like child sex abuse material, content moderation has become “primarily a brand management exercise,” in which companies set their own practices and standards. In particular, she pointed to Musk’s announcement that he would make speech on the platform “as free as reasonably possible” when he bought it. The European Commission is investigating X after its own report showed the platform had cut its content moderation teams by nearly 20% since October 2023. “[Musk's] stated goal in purchasing Twitter was to eliminate content moderation. Now there's no such thing as eliminating content moderation. So what he really meant was, ‘I want to eliminate the content moderation that forces me to behave in a way I don't want to,’” Roberts said. Colorado State University professor Hamed Qahri-Saremi, who researches online social platforms, also cast doubt on how effective the states’ request would be and said he’d be surprised if X agreed. “I don't think these letters would really make much of a change unless there is a more unified approach at the federal level with clear penalties,” he said. “We are far away from being there.”
| 
Comments
Post a Comment