Israeli AI ‘Fake News’ Startup Cyabra Analyzed Biden Harris Announcement: Here’s What They Uncovered

Key Takeawaysdisinforma

• After Joe Biden announced that he would withdraw from the US presidential race on X, thousands of bots commented on his post.
• A post by Kamala Harris thanking the President for his endorsement was also targeted.
• The AI-powered social media campaigns were identified by Israeli anti-disinformation startup Cyabra.

Within hours of Joe Biden announcing his intention to drop out of the US presidential race on X, the social media platform witnessed a wave of deepfake videos and AI-generated bots taking advantage of the situation. 

For example, an analysis by the Israeli cyber intelligence startup Cyabra, which on Tuesday announced its intention to list on the Nasdaq via a SPAC merger, discovered that thousands of fake accounts had commented on Biden and Vice President Kamala Harris’ posts. 

Pro-Trump Bots Spam Biden Harris X Posts

According to Cyabra’s analysis, in the 9 hours following his announcement, 29% of the 6,664 profiles that commented on Biden’s post were fake. Of profiles that commented on Harris’ post, 22% were identified as bots. 

A common characteristic of the fake accounts was the absence of real profile pictures, with many using American flags or images of Donald Trump instead. 

“The majority of these accounts aimed to bolster Trump’s reputation and demonstrate support for him in the upcoming elections,” the report found.

Alongside the large volume of inauthentic comments on Biden and Harris’ posts expressing pro-Trump sentiment, Cyabra identified a separate campaign piggybacking off support for Harris to promote a porn site.

Hashtags Used to Promote Porn Site

After analyzing key phrases and hashtags related to Biden and Harris, Cyabra uncovered a “coordinated and opportunistic campaign” intended to promote the porn site Jerkmate.

The campaign involved nine automated bots sharing identical posts emulating one made by the authentic account of actor Wendell Pierce, which expressed support for Harris. While the text of the fake posts was lifted from Pierce’s word-for-word, they added images extracted from porn videos.

In the space of nine hours, the bot accounts posted nearly 2,000 times, reaching an audience of roughly 200,000 X users.

Cyabra to Become First Public Anti-Disinformation Firm 

Founded in 2017, Cyabra has emerged as one of the leading startups using AI to tackle disinformation. 

The company leverages machine learning to analyze social media and online content, employing natural language processing and sentiment analysis to identify malicious narratives. Key features of the platform include bot detection, influence mapping, real-time monitoring and automated threat assessment.

Leading a cohort that includes Blackbird, Defudger, NewsCheck and Primas, following its anticipated Nasdaq listing, Cyabra will become the first publicly traded company solely dedicated to combating fake news.

Explaining the decision to go public in a blog post , CEO Dan Brahmy said: 

“[AI] is being exploited by bad actors to automate the creation of thousands of fake social media accounts, amplify false narratives, and manipulate public opinion on a massive scale […] Going public gives Cyabra the public and financial platform to take the fight to the bad guys.”