On Thursday, Parag Agrawal, CTO at Twitter stated that the company challenges millions of Twitter accounts every week to check whether a real human is running an account or those accounts are entirely automated.
Twitter CTO was speaking at the Collision from Home tech conference and stated that there are several categories of fake accounts that bad actors can use. He said that some of the fake accounts are entirely automated while some accounts are bots with real humans in the loop.
There are other fake accounts that are operated by humans and attempt to manipulate chats as well as accounts that get compromised and then bad actors use them. Agarwal also revealed that the company suspends millions of Twitter accounts each month before a user even sees those accounts on their Twitter timelines or in the search results.
Twitter applies AI technology to monitor the behaviors of several accounts and detect patterns to determine any abnormal behavior. If the company detects anomalies that cannot be explained by normal human usage, Twitter knows whether to investigate particular accounts or not.
According to Agrawal, figuring out whether an account is fake or not, it is not always easy for the company as there are a lot of accounts with no profile image and a small number of posts that seem like bots. However, those accounts are actually used by real humans. The company also allows users of its platform to use pseudonyms.
Still, Agrawal states that those Twitter accounts that seem to be fake are not potentially the most problematic ones. According to him, the most dangerous fake Twitter account seems to be real on the surface. This makes it difficult for the company to determine whether an account is fake or not.
Post a Comment