FCC Commissioner Warns AI-Generated Robocalls Could Disrupt Information Ahead of US Midterms

FCC Commissioner Anna Gomez, a Democrat at the US Federal Communications Commission, has called on the agency to reject permission for robocalls generated by artificial intelligence (AI) in the run-up to the midterm elections in early November.

According to Reuters and AP, Gomez on Wednesday spoke out against a petition from Club for Growth, a national conservative fiscal policy organization, seeking to allow the use of AI-generated robocalls.

The petition, filed on August 31, asks the FCC to let political groups make automated calls to the public without prior consent. The group pledged to limit the number of calls so that no person receives more than three in 30 days, and to offer an opt-out option.

The FCC opened a public comment period on September 4, with a Monday deadline for submissions. Replies to those comments are expected by October 19, and the FCC has not yet decided whether to grant the request.

Robocalls have been used in political campaigns for years. The FCC bans such calls to mobile phones and pagers without the recipient's prior express consent, but still permits them to landlines.

The concern about AI is that the technology can sound like a real human voice and makes impersonation easier. Amid a rise in scams using the technology, the FCC issued a ban in February 2024, reasoning that "this action makes it illegal to use AI to generate voices in robocalls."

"If the FCC allows a dark-money group to flood Americans' phones with AI-generated robocalls they never agreed to receive, that would cause chaos close to Election Day," Gomez told Reuters on Wednesday.

Speaking to AP, she said removing the FCC's consent requirement could lead to a surge in unwanted automated calls and misinformation. A representative for Gomez did not respond to Al Jazeera's request for comment.

Those concerns are not unfounded. Before the 2024 primary election in New Hampshire, voters received a robocall using an AI-generated voice mimicking incumbent President Joe Biden, urging them not to vote.

The FCC said political consultant Steve Kramer, who worked for little-known candidate Dean Phillips, was responsible and could face a $6 million fine. He was later acquitted of voter suppression charges.

At the state level, Louisiana, Tennessee and Vermont passed laws this year requiring disclosure for AI-generated campaign calls. More broadly, 25 states have enacted laws governing the use of AI-generated material in elections since 2019.

Growing concerns about AI in elections

Concerns about AI robocalls reflect broader anxiety about AI ahead of the midterms. According to a new Reuters/Ipsos poll, 57% of registered voters said President Donald Trump's administration has not taken the risks of AI seriously enough, while 54% said the same about congressional leadership.

Late last month, the White House summoned leaders of top AI companies, including Anthropic, OpenAI, Meta and Google, to sign a voluntary agreement under which the companies agreed to add safeguards around the development of the most advanced AI systems. The voluntary nature of the agreement has not satisfied critics.

Tom Wheeler, a visiting fellow in governance studies at the Brookings Institution's Center for Technology Innovation, wrote of the decision: "The White House's superintelligence agreement is only 'morally binding,' allowing AI companies to design their own controls, choose their own evaluators and determine for themselves whether their processes were followed." He argued that history shows voluntary commitments from these companies are often unreliable, as Facebook, Google and Twitter have all withdrawn or violated their own previously set standards.