Jump to content

Slowed video shows Nancy Pelosi drunk

From bullshitopedia
Slowed video shows Nancy Pelosi drunk
MANIPULATED MEDIA
"A video circulating on social media in May 2019 shows US House Speaker Nancy Pelosi drunk or impaired, slurring her words during a public speech."
Domain Technology & Media
Tactic Manipulation
False Context
Era Current
First seen 2019
Origin A Facebook page that posted a doctored clip of a May 22, 2019, speech at a Center for American Progress event
Last reviewed 2026-10-26
Further reading: books on Technology & Media

The claim that a slowed video shows Nancy Pelosi drunk arose from a doctored clip that spread widely on social media in May 2019. The video showed the then-Speaker of the US House of Representatives speaking at a conference, apparently slurring her words. In fact the original footage had been slowed down to about 75 percent of its normal speed, and the pitch of her voice had been adjusted so that the slowing was not obvious. The original recording shows her speaking normally. The episode became a landmark case in debates about manipulated media because it showed that crude editing, not sophisticated artificial intelligence, could be enough to deceive millions of people, and because of how social media platforms chose to respond.

The claim

The edited video showed Pelosi speaking at a Center for American Progress event in Washington on May 22, 2019. Because the footage was slowed, her speech sounded halting and slurred, and her movements appeared sluggish. Posts sharing the clip suggested or stated that she was drunk, ill or mentally impaired. The claim fit a broader effort by some of her political opponents to question her fitness for office.

Origin

The video was posted on a Facebook page called Politics WatchDog and was quickly shared on other pages and platforms. It was viewed millions of times within days. Some prominent political figures shared or commented on it, and Rudy Giuliani, then the president's personal lawyer, shared a version before deleting his post.

The same week, President Donald Trump posted a different video on Twitter, a compilation from a Fox Business segment that edited together moments in which Pelosi paused or stumbled over words during a news conference. That video was not slowed down, but it was selectively edited to create a misleading impression.

What the evidence shows

Analysis by journalists and digital forensics experts quickly showed that the viral clip had been manipulated. Comparing the edited video with the original recording, published by C-SPAN and other outlets, revealed:

  • The edited clip was slowed to roughly 75 percent of its original speed.
  • The audio pitch had been altered to compensate, so Pelosi's voice did not sound unnaturally deep, as it normally would when footage is slowed.
  • In the original, Pelosi spoke at a normal pace and without slurring.

The Washington Post and other outlets published side-by-side comparisons showing the difference. There was no evidence that Pelosi had been impaired.

Platform responses

The platforms' responses became a significant part of the story:

  • YouTube removed the video, saying it violated its policies.
  • Facebook declined to remove it. The company said it did not have a policy requiring that information posted on its platform be true. Instead, after fact-checkers rated it false, Facebook reduced its distribution in users' news feeds and displayed notices linking to fact-checks when people tried to share it.
  • Twitter did not remove it at the time.

Pelosi criticized Facebook, saying the company had been a willing enabler of false information. The debate led Facebook and other platforms to develop more explicit policies on manipulated media, although those policies have generally focused on content created with artificial intelligence and have often excluded simpler edits.

Why it matters

The Pelosi video is often cited as an example of a "cheapfake" or "shallowfake," a term popularized by researchers Britt Paris and Joan Donovan in a 2019 report. Unlike a deepfake, which uses artificial intelligence to fabricate images or sound, a cheapfake uses simple, widely available editing techniques such as slowing, speeding up, cropping or splicing footage.

The case showed several things:

  • Simple edits can be highly effective. Slowing down a clip required no special technology, yet it deceived many viewers.
  • Real footage can be made to lie. Because the video was based on genuine footage of a real event, it seemed authentic.
  • Existing beliefs shape perception. People who already disliked Pelosi or doubted her fitness were more inclined to accept the video without question.
  • Corrections struggle to catch up. Even after fact-checks, the video continued to circulate, and similar edited videos of Pelosi appeared in later years.

Lessons for readers

  • Find the original footage. A clip of a public event can usually be compared with the full recording from a broadcaster or the event organizer.
  • Notice signs of manipulation, such as unusual pacing, sluggish movements or audio that does not quite match the video.
  • Ask who posted it first. Anonymous or partisan pages are common sources of manipulated clips.
  • Be especially cautious with videos that confirm what you already believe about someone.
  • Manipulated media does not require advanced technology. Many effective fakes are made with basic editing tools.

See also

Sources

  • Drew Harwell, "Faked Pelosi videos, slowed to make her appear drunk, spread across social media," The Washington Post, May 24, 2019.
  • Britt Paris and Joan Donovan, Deepfakes and Cheap Fakes: The Manipulation of Audio and Visual Evidence (Data & Society, 2019).
  • C-SPAN, recording of Nancy Pelosi's remarks at the Center for American Progress Ideas Conference, May 22, 2019.
  • Statements by Facebook and YouTube on the Pelosi video, May 2019, as reported by multiple outlets.