AI Election Deception Puts Voters in the Crosshairs

Artificial intelligence can help attackers impersonate candidates and election officials, and give dishonest actors a way to dismiss authentic evidence as fake, Bobby Ford, chief strategy officer at Doppel, warned. An attacker doesn’t need to change votes to damage confidence in an election. Confusion alone can be enough. Securing election systems, Ford wrote, doesn’t necessarily protect voters from manipulation.

Confusion as a weapon

In 2024, Iranian cyber actors stole private campaign material and tried to distribute it to influence the election. U.S. intelligence agencies attributed a fabricated video showing ballots being destroyed in Pennsylvania to Russian actors. Ford said both tactics sought to exploit divisions and undermine confidence. AI can make deception cheaper, faster and easier to personalize, while helping dishonest actors claim real evidence came from AI.

The targets aren’t only voters. Social engineering exploits how people think, whom they trust and what prompts them to act. During elections, attackers can lean on strong emotions, familiar public figures and a steady stream of consequential information. Ford described hypothetical calls that sound like a candidate asking for an urgent donation, texts claiming a polling place has changed, and videos appearing to show an election official admitting misconduct. AI can generate convincing variations at scale.

Campaign staff can also face attacks. Someone impersonating a colleague could use a cloned voice to request an authentication code. If the deception succeeds, attackers could expose private communications, leak them, edit them selectively or use them in a misleading narrative. An attack might begin with a social media post, continue with a fake website and end with a text. Repetition isn’t verification. Multiple accounts sharing a clip may all rely on one unverified source.

When real evidence gets called fake

Legal scholars Bobby Chesney and Danielle Citron call the danger that deepfake awareness can help people dismiss genuine evidence the “liar’s dividend.” As deepfakes become more familiar, dishonest actors can more plausibly claim authentic recordings, legitimate reporting or other evidence came from AI or were manipulated by it. Ford cautioned that knowing something can be faked doesn’t establish that a particular item is false. Suspicion can’t verify anything.

Ford advised voters to pause when a message urges action, especially if it triggers urgency, outrage or fear. Before donating, sharing or changing voting plans, they should verify requests through a separate channel. They should check registration, polling places and deadlines directly with state or local election offices, and use a campaign’s known website rather than an unsolicited link. An official campaign statement establishes what the campaign says; independent reporting can help assess the claim. Ford also urged voters to say, “I don't know yet,” and wait for evidence rather than immediately judging a viral post.

The institutional response

Ford said campaigns should secure their accounts, preserve original communications and respond quickly to impersonation attempts. Platforms, he said, need effective reporting and escalation processes for deceptive election content. Voters still have to navigate a fast-moving stream of claims while campaigns and platforms manage the accounts and channels attackers may exploit.

Defending Digital Campaigns, a nonprofit organization supported by Ford's organization as a technology partner, provides eligible campaigns with cybersecurity resources. AI can also help defenders identify malicious infrastructure, connect coordinated activity across channels and disrupt campaigns before they reach more people. The article identifies these measures, but offers no electoral or legislative fix for the trust problem.

Technology alone won’t resolve it, Ford wrote. Voters need reliable places to verify what campaigns and election officials have said, and impersonation should be addressed quickly. The danger, he said, isn’t only that AI may persuade people to believe something false. Repeated manipulation may also make them question whether anything is trustworthy. Election security, Ford wrote, includes protecting vote-counting systems and voters’ ability to know what is real.