The following post is an audio file and supplemental transcription.
This post is intended for readers and listeners who may not be politically adept or have a strong grasp on advancing technology and policies surrounding it.
As of October 8, 2026, the Federal Communications Commission is considering an emergency request to loosen rules on political robocalls ahead of the 2026 midterm elections, including those created with AI-generated voices, according to a report from the Associated Press.
The request has been forwarded for public comment, and we can expect a ruling late this month.
There are incredibly concerning implications for AI-generated robocalls in our elections, if the emergency request is granted.
My name is Paul Pascual. I am a student for MCO 427: Misinformation and Society. The following audio and supplemental transcription will demonstrate the evolution of robocalls in political campaigning and how the current literature of misinformation intervention leaves much to be desired, but nevertheless important, in relation to audio technology and direct wireless calling.
In past elections, political robocalls have misinformed voters. In 2020, Detroit residents received a robocall claiming if they voted by mail, they could face arrest and debt collection, according to AP.
And in 2024, an AI-generated voice mimicking President Joe Biden told New Hampshire voters to forgo voting in the primary and save their vote for the general election.
The actors of the robocalls have been subsequently criminally charged and/or fined. However, with this emergency request to the FCC, campaigns and political groups could potentially have the strongest direct campaigning power. And as technology improves, these groups could potentially employ live AI-driven conversations, Ali Swenson of AP wrote.
We have already seen AI-generated political ads, some of which misrepresent candidates. In early September, the campaign for Republican Ken Paxton for Senate released an AI-generated video using Democratic James Talrico’s image and voice.
Ken Paxton is the Republican U.S. senatorial nominee for Texas. Paxton’s opponent is James Talrico, the Democratic nominee.
It’s clear within our country’s extreme political polarization, campaigns and groups are willing to mischaracterize and misrepresent opposing nominees with emerging AI technology, as represented through Paxton’s campaign.
And with this medium of communication, it would be difficult for listeners to distinguish AI-generated or AI-assisted voices and human voices. According to a study published in PLOS One, AI-assisted voice cloning was indistinguishable from human voices with entirely AI-generated voices lagging shortly behind.
In an article published by the Washington Post (Module 3), people tend to believe anything they hear, and the messaging is strengthened when it’s repeated. With the increasingly more human-sounding AI voices and the efficiency of targeted robocalls, with more callers becoming available, misinformation and misrepresentation could potentially become increasingly more relevant and easier to disseminate.
Automated calls reduce the noticeable characteristics of synthetic activity we see in digital media, especially with AI voices. The skill of lateral reading, a foundational skill in combatting misinformation (Module 6), may become harder to do in synchronous communication through a phone call as opposed to asynchronous content on social media and in articles. Additionally, the barrier between the information received through a call and the verification of that information becomes greater as the content leaves the digital space.
This raises a lot of challenges for misinformation intervention and other employable skills to combat misinformation.
Debunking, or fact-checking, information becomes more difficult as misinformation and misrepresentation becomes disseminated directly in addition to the time it takes to combat that information.
Prebunking; however, may have some weight despite the loss of the visual characteristics and context. Prebunking challenges the techniques used in misinformed and intentionally misrepresented communication like high emotional appeal, polarization and manipulating context. By listeners recognizing these technique characteristics, AI-automated robocalls may lose their efficiency.
Nevertheless, the advancing request to the FCC is concerning, and the consequences of the request, if passed, are to be discovered. However, by uncovering the new medium high-volume political campaigning may touch, we become increasingly more aware of how technology and misinformation advances into our lives.
