Skip to content
Search

Latest Stories

Follow Us:
Top Stories

Overcoming AI voice cloning attacks on election integrity

Computer image of a person speaking
ArtemisDiana/Getty Images

Levine is an election integrity and management consultant who works to ensure that eligible voters can vote, free and fair elections are perceived as legitimate, and election processes are properly administered and secured.

Imagine it’s Election Day. You’re getting ready to go vote when you receive a call from a public official telling you to vote at an early voting location rather than your Election Day polling site. So, you go there only to discover it’s closed. Turns out that the call wasn’t from the public official but from a replica created by voice cloning technology.

That might sound like something out of a sci-fi movie, but many New Hampshire voters experienced something like it two days before the 2024 presidential primary. They received robocalls featuring a deepfake simulating the voice of President Joe Biden that discouraged them from participating in the primary.


To be sure, there’s no indication that the fake Biden robocalls had a discernible impact on the New Hampshire primary, but the incident is a stark reminder of the growing threat posed by tactics like this, which are increasingly being used by malign actors to target elections not only in the U.S. but in Slovakia, Argentina and elsewhere.

As artificial intelligence tools become more accessible and affordable, deepfake attacks (of which voice cloning is only one example) are becoming more frequent. How can voters protect themselves from similar efforts to ensure that they make informed decisions for the November general election? Here are a few tips:

1. Avoid answering calls from unknown numbers: Picking up a call from an unknown number increases the likelihood of falling for a scam. Additionally, if you answer a call from an unknown number and speak, a scammer can record your voice and use it to create cloned scam calls to trick your family members and friends.

2. Verify the caller’s identity: If you do answer a call that raises suspicion, take steps to verify the caller’s identity. Several New Hampshire voters did this after receiving the Biden robocall and were able to confirm that the voice was fake. Try to contact the person (or their campaign) through an alternative channel to confirm that the call was actually from the person/organization it purported to be from.

3. Report potential voice cloning: If you may have received an AI voice scam call, contact the appropriate authorities so they can use their expertise to investigate further. This can help address your scam, as well as others, and deter similar future behavior. After New Hampshire voters alerted law enforcement and their attorney general about the robocall that used AI to impersonate Biden, the alleged culprit was identified and charged with 13 counts of voter suppression, a felony, and 13 counts of impersonating a candidate, a misdemeanor. He also faces a proposed $6 million fine from the Federal Communications Commission.

4. Educate yourself: Knowledge is your best defense against emerging threats. Take the time to educate yourself and those around you about the dangers of voice cloning. Be skeptical of unsolicited calls, especially if they involve urgent requests that offer suspicious information or try to get you to engage in behavior that sounds “off” (like sending gift cards to supposed relatives of friends).

5. Rely on trusted sources: Our information ecosystem is awash in lies and inaccurate information, but at least in the elections space we know whom to seek out for accurate information about the administration of elections: state and local election officials (and those who support their efforts).

6. Make a plan to vote in advance of Election Day: Devising a vote plan allows you to confirm when, where and how you can vote. It also enables you to consider alternatives in case your preferred plan for voting does not work out because of something unforeseen like an illness. Finally, planning makes it less likely that you’ll be tricked by something like a voice cloning attack, even if it appears real.

Voice cloning attacks are part of the “new frontier” in malign efforts to meddle in U.S. elections. By staying informed, establishing safeguards, and remaining skeptical of unexpected communications, voters can increase their chances of thwarting these threats before they cause real damage.


Read More

photograph shows a handheld smartphone displaying the icons of some of the main artificial intelligence based apps

This photograph shows a handheld smartphone displaying the icons of some of the main artificial intelligence based apps, including LLMs, chatbots and generative AI, with logos (from L) of Proton AG's Lumo, Meta AI, Mistral Vibe (formerly Le Chat), xAI's Grok, Microsoft's Copilot, Google's Gemini, Anthropic's Claude, Perplexity, Deepseek, OpenAI's Chat GPT, Google's Notebook LLM and generative AI music app Suno, in Saint-Mande, east of Paris, on July 15, 2026.

Photo by Martin LELIEVRE / AFP via Getty Images)

The Second AI Election: Testing the Safeguards Before November

The 2024 U.S. presidential election was supposed to be the “first AI election.” Experts warned that generative AI could flood voters with deepfakes and fabricated evidence of fraud, U.S. intelligence officials declassified intel reports on foreign influence operations already experimenting with the technology, and leading AI companies pledged to combat deceptive election content. But foreign nations faced significant barriers to deploying AI to influence that election, according to the now-shuttered U.S. Foreign Malign Influence Center.

Two years later, the picture looks very different. AI tools have become far more sophisticated and widely available. Russia, China, and Iran are deploying a wide variety of cutting-edge AI tools in more sophisticated ways in foreign influence operations aimed at the United States and its citizens, among others. Meanwhile, the Trump administration has itself used AI in misleading ways, while at the same time dismantling or defunding the federal and independent bodies that identified and countered election-related influence campaigns. With the potential for AI-driven misinformation campaigns to escalate sharply in this year’s midterms, we tested some of the most popular AI models to learn how they can be exploited to peddle false election narratives. The results are clear: AI companies, lawmakers, and civil society must do more to blunt the threat ahead of elections this November and in 2028 to help ensure free and fair elections. We outline the steps they can take to do so below.

Keep ReadingShow less
Robot Holding a Resume Doing HR Work Vector Illustration

AI résumé screening can filter out qualified candidates before a human ever looks. Here's why transparency and human oversight in hiring matter.

nicoletaionescu/Getty Images

When AI Screens the Resume, Who Gets a Fair Shot at Work?

A job seeker can spend hours tailoring a résumé, checking qualifications and writing a thoughtful application, believing the next step will be a person deciding whether to schedule an interview.

Increasingly, that assumption may be wrong.

Keep ReadingShow less
President Trump answers questions from reporters.

U.S. President Donald Trump answers questions from reporters before boarding Air Force One on August 21, 2026 in Joint Base Andrews, Maryland.

(Photo by Win McNamee/Getty Images)

Trump’s White House Press Ban Violates 1st Amendment

The principle is simple. Just like politicians are not supposed to pick their voters, government officials don’t get to pick the reporters who cover them.

If they did, the First Amendment would be meaningless. and we would no longer live in anything resembling a constitutional republic. That is why President Trump’s September 18 announcement that he is banning reporters from CNN, MSNow, and Politico from the White House is so dangerous.

Keep ReadingShow less
A person interacts with an AI chatbot on a smartphone. The screen displays a conversation interface with the text "Hi, may I help you?" suggesting the use of an AI assistant app. A laptop is visible in the background, emphasizing a tech-savvy environment.

A 2026 study found AI beat elite debaters, until it was slowed to human speed. Why persuasion throughput matters and how to make chatbot tempo transparent.

Chong Kee Siong/Getty Images

Political Chatbots Should Not Be Allowed To Set the Pace

A political chatbot can answer before a citizen has finished reading its previous message. It can add statistics, historical examples, causal claims, and links at a rate no human campaign volunteer could match. That does not necessarily make its case wiser. It changes the contest by allowing one speaker to control the tempo.

A June 2026 preprint offers unusually direct evidence of this problem. Across four preregistered experiments involving 18,978 conversations from 6,923 people, frontier AI systems were more persuasive than laypeople, tournament winners, professional canvassers and elite competitive debaters. The human experts were not casual opponents: the study included world and continental champions, paid preparation, live practice and substantial performance incentives.

Keep ReadingShow less