An unpublished paper reported on by The Guardian has found that bots were to blame for a quarter of posts perpetuating climate change denial on Twitter in 2017. The findings from researchers at Brown University further demonstrate the impact that damaging fake news can have on the public’s perception, shaping elections and slowing change.
Companies are able to simulate human users on social media by crafting computer programs that appear to behave in a similar way. They’re commonly used to increase online engagement, boost followers and spread information, whether it's true or not.
Researchers at Brown University analyzed 6.5 million tweets in a draft study that has not yet been published but was reported on by The Guardian. The analyses found that when President Donald Trump announced that the United States was backing out of the Paris climate accord, bot-generated tweets accounted for 25 percent of the messages going out on Twitter about climate change denial, with many tweets applauding the President’s decision. This falsified the impression that climate change denial was a more widespread belief at that time than it necessarily was.
Bots are able to independently post tweets once they have been set up, but there must be a programmer to get them going in the first place. The investigation by Brown has not identified who or how many people are responsible for this widespread misinformation.
To conduct the study, the team borrowed a Botometer from Indiana University that can establish the likelihood of a tweet having been written by a human or a bot. Most of the tweets were relating to “bogus” climate science and rejecting claims that climate change was a pressing crisis and the Botometer categorized these tweets by subject.
It found that bots were responsible for 38 percent of tweets about "fake science" and 28 percent of tweets about the oil company Exxon. The bias of bot use in this type of misinformation became more evident when they analyzed tweets that advocated that climate change was a pressing issue, finding only 5 percent of these could be attributed to bots.
During the days directly surrounding President Trump's announcement that the US was leaving the Paris agreement, there was a general increase in the number of posts about climate change. This included the number of posts by bots, which rose from hundreds per day to over 25,000 per day. Though the Brown University study was unable to identify who set up the bots, the draft paper’s findings suggest there could be a significant impact of mechanized bots in spreading denialist messages about climate change.
The paper sheds light on how so many could still believe climate change isn’t real despite the overwhelming scientific evidence, with lead author of study reportedly telling The Guardian that his team were “always kind of wondering why there’s persistent levels of denial about something that the science is more or less settled on”.
The facts have long-contradicted climate change deniers, fortunately followers of the belief-based conspiracy theory in search of inspiration need look no further than the President’s colorful Twitter history: