Skip to content
EN

Lesson 3 of 8 · 6 min · +100 XP

How can you spot a deepfake, especially in an election?

Election deepfakes are AI-made videos, images or voice clips that show candidates saying or doing things they never did, often released just before a vote. To spot a deepfake, pause, check who posted it first, look for the clip on trusted news sites or fact-checkers, and don't share until it's confirmed. Report it to the platform.

▶ Try it

Why it matters

AI can fake a candidate's voice or face in minutes. Timed right before a vote, a fake clip can spread faster than any correction. The real defence isn't a perfect detector, it's people who pause before they share.

In 6 minutes you'll be able to

  • Fake voices and last-minute clips are the most common election tricks.
  • Check the first source, trusted news and fact-checkers, not just a detector.
  • Voting info only from the official election website. Report fakes.

What you'll practise

  1. Four election tricks

    Deepfakes in politics usually try one of these.

    🎧 Fake voice
    A cloned voice of a candidate in a robocall or "leaked" audio. Audio is easy to fake and hard to check.
    ⏰ Last-minute timing
    Released just before voting, when there's no time for a correction to catch up.
    🗳️ Fake voting info
    "Polls are closed", "vote by text", "vote on Wednesday". Designed to stop people voting.
    🙈 The liar's dividend
    Once fakes exist, people caught on real recordings can simply claim "it's a deepfake".

  2. The viral clip

    This post is spreading fast. Tap every warning sign.

    TruthDrop_2026
    🚨 LEAKED AUDIO 🚨 Candidate Ellis admits the vote is rigged. Don't bother voting tomorrow, it's already decided.
    Mainstream media won't show you this!
    Share before it's taken down!!
    Polls open at 7 am.

  3. What would you do?

    Real situations, sensible moves.

    1. Your uncle forwards a voice message of the mayor saying the vote has moved to Wednesday. What now?

      • Check the official election website, then tell him kindly (Answer)
      • Forward it to the family group to be safe

    2. A politician says a real, verified video of them is "just a deepfake". What's going on?

      • Possibly the liar's dividend: blaming AI to dodge real evidence (Answer)
      • Any video can be fake, so all videos are worthless

    3. An online detector says a clip is "97% real". Settled?

      • No, detectors make mistakes, so also check the source (Answer)
      • Yes, 97% is basically certain

  4. Your two-minute check

    A shocking political clip lands in your feed. Put the steps in order.

    1. Pause: strong feelings plus a deadline mean slow down
    2. Check who posted it first and when
    3. Search trusted news sites and fact-checkers for the clip
    4. For voting dates or rules, check the official election website
    5. Don't share it; report it to the platform if it's fake

Remember

Shocking clip right before a vote? Pause, check, and don't share until it's confirmed.

Key facts

  • In January 2024, robocalls with an AI-cloned voice of President Biden told New Hampshire voters not to vote in the primary. The FCC later fined the consultant behind them $6 million. · Source: US Federal Communications Commission, 2024
  • In February 2024, the FCC ruled that AI-generated voices in robocalls count as "artificial" under the Telephone Consumer Protection Act, making such calls illegal without prior consent. · Source: US Federal Communications Commission, Declaratory Ruling, February 2024
  • Under the EU AI Act, people who publish deepfakes must disclose that the content was artificially generated or manipulated, from 2 August 2026. · Source: Regulation (EU) 2024/1689 (AI Act), Article 50
  • Researchers call it the "liar's dividend" when the existence of deepfakes lets people dismiss real evidence as fake. · Source: Chesney & Citron, Deep Fakes: A Looming Challenge, California Law Review, 2019

Updated:

Questions people ask

Are political deepfakes illegal?

It depends on where you are and what the fake does. Some uses are already illegal, like fake AI-voice robocalls in the US, and the EU requires deepfakes to be labelled. Many places are still writing rules.

How can you tell if a video is a deepfake?

Look for odd lips, lighting or voices, but don't rely on them. The best checks are who posted it first, whether trusted outlets confirm it, and a reverse image search of key frames.

Where do I report an election deepfake?

Use the platform's report option for false or manipulated media. If it spreads false voting information, also tell your official election authority.

What is the liar's dividend?

It's when people caught on real recordings claim they're deepfakes. Because fakes exist, real evidence becomes easier to deny.

Related terms