FCC Democrat Warns AI Robocalls Could Flood Voters Ahead of Midterms

A Democratic member of the U.S. Federal Communications Commission is raising alarms over a conservative group's request to unleash AI-generated robocalls on American voters just weeks before the midterm elections. Commissioner Anna Gomez said Wednesday that approving the petition would invite chaos at a moment when voters are already on edge about the integrity of election information.

At issue is a petition filed August 31 by Club for Growth, a national conservative fiscal policy organization, asking the FCC to let political groups place automated phone calls to people without first obtaining their permission. The group has said it would cap its outreach, promising that no individual would receive more than three calls in a 30-day window and that recipients would be given a way to opt out.

The FCC opened a public comment period on September 4, with the deadline for submissions falling on Monday. Replies to those comments are due October 19, and the commission has not yet indicated whether it will grant the request. The timeline means a decision could land in the final stretch before voters head to the polls in early November.

Gomez, who has been an outspoken voice on the five-member commission during President Donald Trump's administration, told Reuters that allowing a dark money group to swamp Americans' phones with AI-generated robocalls they never consented to would bring disorder dangerously close to Election Day. She made similar remarks to The Associated Press, warning that dropping the FCC's consent requirement could produce a spike in unwanted calls and a corresponding rise in false or misleading information. Representatives for Gomez did not respond to a request for comment from Al Jazeera.

The regulatory backdrop is important. Robocalls have been a fixture of political campaigning for years, and the FCC already bars such calls to cellphones and pagers unless the recipient has given prior express consent. Landlines are treated differently, and political calls to those numbers are generally permitted. What has changed is the technology behind the voice on the other end of the line.

Artificial intelligence tools can now mimic a real person convincingly, making it far easier to impersonate candidates, officials, or ordinary constituents. Citing a surge in scams built on that capability, the FCC moved in February 2024 to outlaw AI-generated voices in robocalls, stating at the time that the action made using AI to generate the voice in these calls itself illegal.

The commissioner's concerns are not theoretical. Ahead of the 2024 New Hampshire primary, voters received a robocall featuring an AI-generated clone of President Joe Biden's voice, the Democrat then in office, urging them not to cast a ballot. The FCC identified political consultant Steve Kramer, who was working for long-shot primary challenger Dean Phillips, as the person responsible and said he could face $6 million in fines. Kramer was later acquitted of voter suppression charges.

States have moved on their own as federal rulemaking grinds forward. Louisiana, Tennessee, and Vermont enacted laws this year requiring disclosures when campaign phone calls are generated by AI. Since 2019, 25 states have passed legislation addressing the use of AI-generated materials in elections, a patchwork that leaves voters in other states with varying levels of protection.

Broader anxiety about AI's role in elections appears to be mounting. A new Reuters/Ipsos poll found that 57 percent of registered voters believe the Trump administration has not taken the risks posed by AI seriously enough, while 54 percent said the same about congressional leadership. The findings suggest the issue resonates beyond party lines as the midterms approach.

The White House sought to address those worries late last month by convening leaders from major AI companies, including Anthropic, OpenAI, Meta, and Google, to sign a voluntary pact committing them to stronger safeguards around their most advanced systems. Critics have noted that the agreement is not compulsory, leaving firms to police themselves.

Tom Wheeler, a visiting fellow in governance studies at the Brookings Institution's Center for Technology Innovation, questioned whether such pledges carry weight. He wrote that the White House accord is morally binding only, allowing AI companies to design their own controls, choose their own evaluators, and decide whether their own procedures are being followed. He added that history suggests voluntary commitments from these companies are unreliable, pointing to Facebook, Google, and Twitter, which he said have all walked back or violated previous self-imposed standards.

For Gomez, the immediate question before the FCC is narrower but no less consequential: whether political operators may use synthetic voices to reach voters who never agreed to be called. Her answer, delivered with the election calendar ticking down, is an unequivocal no. Whether the commission's majority agrees may not be known until well after the comment period closes later this month.