15 thousand bots are deceiving users on Twitter

New research from Duo Security has shed light on a notorious Twitter phenomenon that involves scamming naive users out of money under the guise of giving away free cryptocurrency. After analyzing 88 million Twitter accounts, experts discovered 15,000 fraudulent bots.

The search and study of Twitter bots was carried out using specially trained classifier bots. By analyzing only the last 200 tweets from each account, the classifiers found 15,000 active bots spreading false information and posing as well-known personalities in the crypto space.

“Users tend to trust a tweet depending on the number of shares or likes. Fraudsters know this and have designed bots to meet these criteria,” said Olabode Anis, one of the Duo researchers, in a press release.

Famous individuals who have fallen victim to scammers who misappropriate other people’s names have long added “I am not giving away cryptocurrency” to their Twitter handles to warn other users. The founder of the Ethereum network, Vitalik Buterin, was one of the first to do this, and others followed.

For active users of “crypto Twitter,” bots have become part of the environment, as their fake promotions are everywhere. Despite the huge number of these bots, they are actively working to ensure that they are not banned. “Because bots are clever at camouflaging themselves to avoid detection, it was necessary to analyze each account holistically, including the metadata associated with the content. For example, bot accounts typically post in a series at short intervals, resulting in a very short average time between tweets. Identifying such features could help expose other malicious spam networks and bots, say the experts, who are preparing to present their findings today at the 2018 Black Hat USA security conference in Las Vegas. Commenting on Duo Security's findings, Twitter said it was "aware" of the problem and that "spam and some automated systems are prohibited by Twitter rules." In many cases, automated systems do not detect spam content,” adding in his justification that “the number of such scam accounts on Twitter is less than 5%.”

According to https://cointelegraph.com