In the past 11 months, someone has created thousands of fake automated Twitter accounts, perhaps hundreds of thousands, to praise former President Trump.

In addition to posting praise for him, the fake accounts ridiculed his critics from both sides and attacked Nikki Haley, the former South Carolina governor and United Nations ambassador who is challenging her former boss for the 2024 Republican presidential nomination.

When it came to Florida Gov. Ron DeSantis aggressively suggested to the bots that he wouldn’t be able to beat Trump, but would make a great running mate.

As Republican voters rate their 2024 candidates, whoever created the bot network tries to put a thumbs up, using online manipulation techniques pioneered by the Kremlin to sway Twitter’s conversation about candidates, while trying to manipulate the algorithms of use the digital platform to maximize their reach.

The sprawling bot network was discovered by researchers at Cyabra, an Israeli technology company that shared its findings with the Associated Press. While the identities of those behind the network of fake accounts are unknown, Cyabra’s analysts determined it was likely made in the US.

To identify a bot, researchers look for patterns in an account’s profile, list of followers, and the content it posts. Human users usually post on a variety of topics, with a mix of original and reposted material, but bots often post repetitive content on the same topics.

Twitter bots helped build the cult of Elon Musk and Tesla. But who makes them?

That was true for many of the bots identified by Cyabra.

One account will say: Biden is trying to take our guns; Trump was the best, and another will say January 6 was a lie and Trump was innocent,” said Jules Gross, the Cyabra engineer who first discovered the network. “Those voices are not people. In the interest of democracy, I want people to know that this is happening.

Bots, as they are commonly referred to, are fake automated accounts that became notorious after Russia used them to interfere in the 2016 presidential election. While major tech companies have improved their detection of fake accounts, the network identified by Cyabra shows that they remain a powerful force in shaping online political discourse.

The new pro-Trump network is actually made up of three different networks of Twitter accounts, all of which were created in bulk during April, October, and November 2022. In all, researchers believe hundreds of thousands of accounts could be involved.

DeSantis leads Trump by a wide margin in California at the start of the primary race, polls show

The accounts all contained personal photos of the alleged account holder and a name. Some accounts posted their own content, often in response to real users, while others reposted content from real users, further amplifying it.

McConnell… traitor! one of the accounts posted in response to an article in a conservative publication about GOP Senate Majority Leader Mitch McConnell (R-Ky.), one of many Republican critics of Trump targeted by the network.

One way to measure the impact of bots is to measure the percentage of posts on a given topic generated by accounts that appear fake. The percentage for typical online debates is often in the low single digits. Twitter itself has said that less than 5% of its active daily users are fake or spam accounts.

However, when Cyabra researchers investigated negative reports about specific Trump critics, they found much more inauthenticity. For example, almost three quarters of the negative posts about Haley could be traced back to fake accounts.

CPAC shows that the GOP is deeply divided heading into 2024

The network also helped popularize a call for DeSantis to join Trump as his vice presidential running mate, a result that would serve Trump well and allow him to avoid a potentially acrimonious match if DeSantis meets participates in the race.

The same network of accounts shared overwhelmingly positive content about Trump and contributed to a widespread misperception of his support online, researchers found.

Our understanding of what the prevailing Republican sentiment is for 2024 is being manipulated by the prevalence of bots online,” the Cyabra researchers conclude.

The triple network was discovered after Gross analyzed tweets about various national political figures and noticed that many of the accounts posting the content were created on the same day. Most accounts remain active, although they have a relatively modest number of followers.

A message left with a Trump campaign spokesperson was not immediately returned.

Most bots aren’t designed to persuade people, but to amplify some content so more people see it, said Samuel Woolley, a professor and disinformation researcher at the University of Texas whose most recent book focuses on automated propaganda.

When a human user sees a hashtag or piece of content from a bot and reposts it, they’re doing the work of the network for it and also sending a signal to Twitter’s algorithms to further boost the spread of the content.

Bots can also succeed in convincing people that a candidate or idea is more or less popular than he or she actually is, Woolley said. For example, more pro-Trump bots could lead people to exaggerate his overall popularity.

Column: Trump tormentor, whiteboard wizard, it’s the brand that matters in the California Senate race

Bots definitely affect the flow of information, Woolley said. They are built to create the illusion of popularity. Repetition is the nuclear weapon of propaganda and bots are very good at repetition. They’re very good at getting information in front of people.”

Until recently, most bots were easy to spot due to their clumsy spelling or account names containing nonsensical words or long strings of random numbers. As social media platforms got better at detecting these accounts, the bots got more sophisticated.

So-called cyborg accounts are an example: a bot that is periodically taken over by a human user who can post original content and interact with users in human ways, making them much harder to sniff out.

Bots could soon become a lot sneakier thanks to advancements in artificial intelligence. New AI programs can create lifelike profile pictures and posts that sound much more authentic. Bots that sound like a real person and use deepfake video technology can challenge both platforms and users in new ways, said Katie Harbath, a fellow at the Bipartisan Policy Center and a former Facebook director of public policy.

Viral ChatGPT carries propaganda and hacking risks, researchers warn

The platforms have gotten so much better at fighting bots since 2016, Harbath said. “But the types we’re starting to see now, with AI, can create fake people and fake videos.”

These advances in technology are likely to ensure that bots have a long future in US politics as digital foot soldiers in online campaigns, and as potential problems for both voters and candidates trying to defend themselves against anonymous online attacks.

There has never been so much noise online, says Tyler Brown, a political consultant and former digital director for the Republican National Committee. How much of it is malicious or even unintentionally unfactual? It’s easy to imagine that people can manipulate that.