Although certain Twitter bots can be really useful, there is another type that pretends to camouflage itself and impersonate real profiles to manipulate information, influence political issues in hot moments such as the run-up to an election or broadcast hoaxes. Not in vain, the bot issue was the great point of contention that made Elon Musk finally not buy the little bird’s social network despite his interest.
If we look closely, there are certain types of behavior that can lead us to suspect that a profile really is a bot. Next, we are going to see what practically unequivocal signals can lead us to identify those Twitter bots from real profiles.
Very changing profiles
One of the aspects that could lead us to suspect that a Twitter account is really a bot could be the frequency with which it changes its name. It is normal that sometime in life we change our nickname or we want to try to find the closest thing to our real name, but when the changes occur indiscriminately and for no apparent reason to change from “RandomName1” to “RandomName2”, we should suspect.
Mariluz Congosto@congosto1/ I discovered a profile by chance reading the TT. His face seemed strange to me and met the parameters of artificially generated images. His behavior seemed strange to me, drawing my curiosity.
Almost a year after monitoring it -> thread 🧵https://t.co/clXInuU9J4 https://t.co/AX4WsHiGjkSeptember 26, 2022 • 9:15 PM446
49
Another aspect to take into account is the period of activity of these bots that arise for manipulation purposes. We can all be disconnected from social networks for a while and have peaks of activity coinciding with some event or event that has to do with our interests, but these bots have additional suspicions. For example, they are activated at key moments such as the days of reflection before elections or an event that can positively influence the political sign that is intended to be favored or against the electoral rival.
In these peaks of activity, they tend to sneak into the hottest trending topics , to enter directly into the limelight and interact directly or indirectly with real profiles related to those TTs or to confront those who use it to expose their contrary ideas.
Another fact that could be suspicious is the fact that, even in old accounts, we find an unusual number of tweets for an account that is supposedly active at present. This is due to the fact that they delete their past tweets .
A bot is given away by the profile picture
Finally, and this deserves a separate point, since it can be used not only in bots that want to manipulate or misinform by transmitting hoaxes, but also for other types of deception, is the fact that generally these accounts place a profile image generated by Artificial Intelligence .
Despite how much this technique has evolved and the realism it offers, there are still certain signs that we can look for to try to identify if an image is real or not. In addition to the classic search to see if the image is stock or is already being used in other accounts, there are certain patterns when generating the image that can lead us to suspicion.
madrigal@SoyMmadrigalTake a look at the images below and you will see why. Artificially generated images always/very often align features in such a way that eyes and mouth are in the same position
Take a look at https://t.co/Om192JWSCWMarch 31, 2020 • 16:5396
4
Artificially generated images often have their own “canon of beauty”. That is, they usually have their mouths and eyes aligned , presenting images that are too perfect to be believable if you look closely.
If you see a Twitter profile that meets all or most of these signs, be wary. Twitter reveals that it removes more than a million bot and spam accounts daily to reduce their harmful reach on the social network, but there will still be a few million more active.
