Armies of bots battled on Twitter over Chinese spy balloon incident

- Advertisement -


The Chinese spy balloon that floated over Canada and the US in 2023 shortly before it was shot down over South Carolina

The Chinese spy balloon that floated over Canada and the US in 2023 shortly before it was shot down over South Carolina

Joe Granita/ZUMA Press Wire/Shutterstock

Tens of thousands of bots tussled on Twitter to try to shape the debate as a Chinese spy balloon flew over the US and Canada last year, according to an analysis of social media posts.

Kathleen Carley and Lynnette Hui Xian Ng at Carnegie Mellon University in Pennsylvania tracked nearly 1.2 million tweets posted by more than 120,000 users on Twitter – which has since been renamed X – between 31 January and 22 February 2023. All tweets contained the hashtags #chineseballoon and #weatherballoon, discussing the controversial airborne object that the US claimed China had used for spying.

Read more


The spy balloon saga of 2023 inflated US-China political tensions

The tweets were then geolocated using Twitter’s location feature, and checked with an algorithm called BotHunter, which looks for signs that an account isn’t controlled by a human.

“There are lots of different things (identifying a bot) is based off, but examples are whether your messages are being sent out so fast that a human literally can’t type that fast, or if you’re geotagged in London one minute, then in New Zealand when it’s physically impossible for a person to do so,” says Carley.

The researchers found that around 35 per cent of users geotagged as located in the US exhibited bot-like behaviour, while 65 per cent were believed to be human. In China, the proportions were reversed: 64 per cent were bots and 36 per cent were humans.

Sign up to our The Daily newsletter

The latest science news delivered to your inbox, every day.

Sign up to newsletter

Of those accounts purporting to be located in neither country, 42 per cent were bots and 58 per cent were humans. While reliable numbers are hard to come by, previous research suggests that between 10 and 20 per cent of users on Twitter are bots. Bots autonomously carry out tasks such as sending Twitter messages to people on the platform and liking other posts. They are often used to try to influence public opinion.

“You’re seeing more bot activity from the tweets that appear to be coming out of the Chinese community than we are seeing coming out of the American community,” says Carley. The overall proportion of bot accounts was also higher in discussions around the Chinese spy balloon than in other events, according to the researchers.

In one example, a China-based bot posted: “#USA #China #14February (…) One could speculate that the US is using the #ChineseSpyBalloon ‘excuse’ to escalate tensions with #Beijing.. Recall that US airspace is highly controlled and that there are more accurate satellite technologies for spying.”

Read more


China, UK and US are all boosting their spy balloon programs

Neither Carley nor Ng would speculate on who was behind the bots, but Steven Buckley at City University of London says “there is likely going to be a mix of both state and individual actors who are seeking to sway and manipulate public opinion on breaking news events”.

As to whether the bot activity made a difference, Carley says: “The fact that the bots are talking a bit differently to the humans meant what people were reading looked a little bit different, and future conversations look different.” And for that reason, says Buckley, it is important to be “incredibly wary” of content we encounter online – and to assume it might not be posted by a human.

Journal reference

EPJ Data Science DOI: 10.1140/epjds/s13688-023-00440-3

Topics:

  • politics/
  • social media
FacebookTwitterEmailLinkedInPinterestWhatsAppTumblrCopy LinkTelegramRedditMessageShare
- Advertisement -
FacebookTwitterEmailLinkedInPinterestWhatsAppTumblrCopy LinkTelegramRedditMessageShare
error: Content is protected !!
Exit mobile version