Pre-Election Day: Hannity’s Voter Fraud Misinformation — A warning on how misinformation on voter fraud will go viral this week

Avaaz’s misinformation war room is working around the clock to spot the most egregious narratives that could influence the American electorate this week.

I want to share a note from our program lead, Fadi Quran, about an out-of-context video of Biden on voter fraud that has reached approx 17 million views and is gaining momentum without any action from tech companies. Fadi believes this is a harbinger of how misinformation will travel in the coming days and is an important case study to be across (sorry we know it's long but lots of good stats and graphs worth scanning). As Fadi said to me earlier: “This video continuing to go viral right before Election Day with no action from the platforms is ridiculous.”

See below for Fadi’s full memo:

“In our monitoring of disinformation trends, we saw a misleadingly edited, out-of-context video of Biden talking about voter fraud continuing to go viral without action from the major tech companies, garnering approximately 17 million views. This is a flashing red warning sign on how platforms are still unprepared in the face of misinformation, and we believe this is a case study of one big way misinformation will spread this week that we all need to be aware of. Stopping these seeds of misinformation from taking root is essential for protecting voters.

Here’s what we saw happen:

On October 24th Presidential candidate Joe Biden, in a conversation on Pod Save America, was asked about election protection efforts. Part of what he said in his response was: “Secondly, we’re in a situation where we have put together, and you guys did it for our administration — President Obama’s administration before this — we have put together I think the most extensive and inclusive voter fraud organization in the history of American politics.”

In context, as confirmed by fact-checkers on the 29th, it is clear he was discussing the prevention of voter fraud, but a misleading edit of that clip with Biden just saying that sentence has gone viral, with pages and groups with over 5.2 MILLION followers sharing it on Facebook, and the video itself being viewed millions of times across Facebook, Twitter, and YouTube.

This video primes viewers with the claim that a major election fraud effort is underway, and is likely to get a second viral wind in coming days if platforms don’t step in..

It is unclear why Facebook, Twitter and YouTube have not taken any significant measures against it — given that it stokes fear that there is widespread voter fraud and undermines trust not only in the candidate but in the whole electoral process. What is also interesting in this case is how the video is being shared across platforms — using gaps in those platforms’ anti-misinformation policies to amplify the content — something we expect to see on Election Day.

Here’s are the details of what our team found:

1. President Trump, the GOP War Room, and YouTube influencers shared the video on YouTube where it has now garnered at least 800k views. YouTube does not have any fact-checking policies on this front, and the video has no labels.
2. RNC Research also shared the video on Twitter.
3. On Facebook, a quick search shows that articles, links, and claims connected to the video have over 400k estimated interactions — which could translate to, at a minimum, approximately 4 million and at a maximum 12 million estimated views on the misleading claim based on similar examples of misinformation content Avaaz has studied. Just by looking at the YouTube links to the video that were shared on Facebook, you can add other 70k interactions. And then, you look at links to the video on Twitter that were shared on FB — those links also have 27ks of interactions.
4. If you specifically do an image meme search on CrowdTangle for the claim, that’s other 67k interactions. On Facebook only, the actual video in some cases has the “misleading” label (from Erik Trump’s Facebook page where it has over 109K views) — But centrally, it appears Facebook does not label misinformation from candidate-related pages — and here the Team Trump Facebook page shared a video, garnering over 264K views, and the video is not fact-checked or labelled. Moreover, now users are sharing YouTube and Twitter links to the video on Facebook, and those posts are not fact-checked or labelled on the platform at all.

Memes sharing the content:

5. On Twitter overall the video now has 8.4 MILLION Views and retweeted over 6.6 thousand times. As of October 31st, it was still being re-shared, and the story amplified and re-packaged on the platform — for example, Sean Hannity uploaded the video on October 31st, and it has received over 780K views since.

Putting all those numbers together, we’re talking, just in the last week, at least 17 million estimated views of this video and connected misinformation content, which has been fact-checked (see here and here). This estimate was made with the publicly available data.

The misleading claim is also getting shared on Instagram, with a meme posted by the well known Kremlin-connected website Sputnik News, from 2016 acclaim:

This is an interesting case of exactly what may happen on election day and throughout this week. Essentially, a “Voter Fraud” claim or incident gets taken out of context somewhere, it is put up on YouTube or Twitter first in a dangerously misleading way — influencers share it in videos, memes and links on Facebook, millions see it and believe it, and by the time the platforms act (if they even catch ALL the misinformation being shared), the harm is already done.

This is also an interesting case study on the sophisticated way in which disinformation is spread across different platforms, in a way that is hard to stop.

All the platforms do have solutions to stem this harm but are not implementing them. For example, when such content goes viral, they can ensure that they:

a) Demote the post and the accounts of those sharing it to limit their reach, instead of allowing the algorithm to amplify the outrage and confusion and recommend the content in users’ newsfeeds.

b) Coordinate to share fact-checks from independent fact-checkers retroactively, ensuring those who saw it before it was labelled get a clear correction and link to a fact-check. Experts across the field have now called for this, including a consensus of leading researchers on debunking disinformation from MIT, Cambridge, and other institutions.

c) Cooperate more seamlessly across platforms to detect and flag disinformation posts and trends amongst platforms quickly to ensure cross-platform posting does not boost the content — but allows it to be labelled.

Feel free to get in touch with any questions you may have. Our team is monitoring misinformation trends around the clock.