The Online Harassment of Meghan Markle Was a Coordinated Attack

The online hate against Harry and Meghan, the Duke and Duchess of Sussex, was carried out by just a few Twitter accounts in a coordinated attack, according to to the Twitter analytics service Bot Sentinel.

As reported by BuzzFeed News, just 83 Twitter accounts were responsible for approximately 70% of the negative content on the couple. Those 83 accounts, however, did have 187,631 followers. And overall, the report suggested the negative campaign by such few people could actually reach 17 million users.

“There’s no motive,” as Bot Sentinel CEO Christopher Bouzy told BuzzFeed. “Are these people who hate her? Is it racism? Are they trying to hurt [Harry and Meghan’s] credibility? Your guess is as good as ours.”

As well, the people behind those tweets certainly know what they’re doing. As Bouzy noted, “This campaign comes from people who know how to manipulate the algorithms, manipulate Twitter, stay under the wire to avoid detection and suspension. This level of complexity comes from people who know how to do this stuff, who are paid to do this stuff.”

As part of their research, Bot Sentinel analyzed 114,000 tweets using hashtags and keywords related to Harry and Meghan, Duke and Duchess of Sussex. Internal tools were used to categorize tweets based on sentiment and repetitiveness, and the site manually examined tweets for “altered photos, deceptively edited videos, and other content our internal tools couldn’t automatically or accurately analyze.”

Bot Sentinel, which is both a site and a browser plug-in, was started by Bouzy in 2018 to analyze and identify Twitter users who violate the social media site’s terms of service. Bots and people behind targeted negative campaigns are the main concerns. The site has also recently identified “inauthentic” accounts related to Project Veritas, a right-wing activist group accused of releasing a deceptive video created to malign Pfizer employees and the Pfizer COVID-19 vaccine. As well, the site has examined online racial abuse of English footballers and thousands of Brazilian bots that were spreading false news of election fraud during the 2020 U.S. presidential election.

The site claims to be non-partisan. “Our machine learning model and algorithm doesn’t care if an account is a Conservative or Liberal, it wasn’t developed for that,” as their FAQ notes.

If you play around on the site for a bit, you can find some interesting trends among the problematic accounts. Currently, the top hashtag amongst inauthentic Twitter users is #nowplaying, the top two-word phrase is “follow back,” the top mention is YouTube and the top URL (which we won’t link to) appears to be from a Japanese account that has a Tucker Carlson video segment. You can also use the site to search through hundreds of thousands of problematic, disruptive and deactivated accounts.