Robo-COP29: Bots Boosted Propaganda Promoting Petrostate Host

Big Tech let thousands of bot accounts promote Azerbaijan’s propaganda before and during UN climate negotiations again this year. The Climate Action Against Disinformation coalition is releasing a brief on the bots that promoted Azerbaijan and the petrostate’s president, and shouted down critics of the regime across social media. 

Two reports published just ahead of COP29 revealed thousands of accounts were being created and used to serve a public relations function for Azerbaijan, and this new brief adds a third set of findings regarding thousands more posts, largely timed in what appears to be an attempt to offset critical stories- including, ironically, news of the bots. 

“Once again CAAD is forced to do what Big Tech can’t, or won’t, do, and identify content that breaks terms of service and puts the public at risk,” said CAAD research coordinator Sean Buchan. “Despite all the hype about AI or the metaverse, the continued failure to protect from the most basic coordinated inauthentic behaviour is not surprising. It’s instead a painful reminder platforms won’t clean up their act unless we make them.” 

New Findings:

• On Facebook, CAAD found evidence that a post from Abzas media, a critic of the Azerbaijan government, was brigaded by likely hundreds of accounts with only a handful of followers, profile pictures copied from other social media, and an odd post consisting of a letter of the alphabet. 

  • The story being brigaded was about news coverage of bots flooding out #COP29 and #COP29Azerbaijan.

     

• CAAD identified 554 suspicious X/Twitter accounts that were largely created in bursts, created and posting during Azerbaijan waking hours. 

  • 464 of those X/Twitter accounts posted identical text, “#COP29 #COP29Azerbaijan” a total of 5,632 times in the two weeks leading up to COP, the majority of which, 4,333, were quote tweets of the official @COP29_AZ account. 

  • A further 481 were quote tweets of content from Azerbaijan President Ilham Aliyev, and 346 quote tweets of the Assistant to the President. 

  • Two-thirds of the tweets posted the day of and after reporting that Azerbaijan will increase gas production by a third in the next decade, suggesting a tightly coordinated rapid response mechanism.

  • A second flurry of tweeting occurred on November 4th, when three critical reports were published on Azerbaijan’s “dreadful human rights and environmental record,” the country’s crackdown of critics, and the arrest of a climate activist, again demonstrating the bots’ public relations purpose.

     

• On YouTube, two promotional videos from Azerbaijan’s COP29 channel were viewed over 1.5 million times, boosted by advertising and with similar hashtag-promoting comments. One 58-second clip on “inclusion and accessibility at COP29” was viewed over 840,000 times, an impressive showing relative to the 50,000 attendees for whom information about accessibility options at the conference center would be relevant.  

This is not the first time Big Tech has failed to mitigate the damage done by disinformation campaigns, as fake accounts were used to defend the host country last year, when we similarly raised concerns about digital climate disinformation, and analysis of COP27 estimated 6% of the X/Twitter conversation was suspected bots. 

Yet Big Tech remains unwilling or incapable of protecting users from these repeated, predictable, easily identified and clearly coordinated inauthentic attacks, used to suppress criticsm and promote petrostates. 

As a result, CAAD, along with nearly 100 organizations and experts, have issued an open letter calling on governments to hold Big Tech accountable for its role facilitating the spread of Big Oil’s harmful climate disinformation. 

—