Among the followers of most accounts, and not just those of some politicians, there is a mixture of ‘bots’ and real human beings. Here are some tips to distinguish Fake Profiles On Social Networks.
The space of free expression that was originally the social networks has been filled over time with noise, criticisms and arguments.
But not all those who generate that noise are real people, and not all the apparent bots are machines. Most accounts have followers with suspicious traits. We show you some clues that can help you to remove their mask.
Let’s start with the photo
A very common feature is that the profile picture of a fake account shows the default avatar from Facebook, Instagram or Twitter. That is, without a photo.
Or, on the contrary, we can come across very neat images. People who look like models (and they probably are, although they have no idea that their image is being used).
Fake profiles can be distinguished by inconsistencies: if you see a picture of a beautiful blonde girl and you see that the description does not match.
How to check the origin of an image
To find out whether the photo has indeed come from an image bank or has been usurped from any other site, we can use Google’s reverse image search, which will show us where it has been published.
To do this, just right-click on the photo and select “search image in Google“. If you do not get that option directly, open the image in a new tab (right click: open link / image in a new tab) and try again. And voila, maybe Karen99’s photo will appear in 200 results from various sources.
There are also images that may have even been computer generated: If I see a face that does not look like a natural face, it may be made with AI, in which strange faces come out, because they put two faces together, and those profiles are possibly fake.
Tell me your name
There are two possible cases: the name may be strange (for example, one that is clearly foreign) or it may be full of numbers.
When we register in a social network it is possible that our name collides with other existing nicks, so we are usually offered alternatives.
If, for example, Karen is already taken, you may be offered something similar to Karen851995, but it is normal to opt for something simpler or more personal.
It is true that not everyone decides to modify it. There are people who for lack of knowledge or for whatever reason keep it. But such a name is an indication that the account may have been created automatically, since in that case no one bothers to change the default name.
Recently created accounts tend to be suspicious when they are used for a specific purpose. Such as supporting or attacking a political party or a company.
But there is also a market in which old ones are reused, which means that there are many second-hand ones. To see if there is a disparity, you can compare the date of creation with its activity.
If, for example, it dates from 2011 and has very few publications, it may simply be that the account has been abandoned, but if after nine years it has recently started its activity, then it is more suspicious.
First and second hand accounts are marketed. With better or worse quality: with or without photo, with more or less followers… It all depends on what you want to pay.
Imbalance between followers and followed
On the one hand, there is the disparity that a profile can follow many people but no one follows it. Which leads one to think that it is a bot.
But on the other hand, there are accounts that have a very high number of followers despite having practically no activity – or very recent activity.
This is due to the second-hand market, which brings us to the next point.
I change my name (again)
In this trade there is the consequent change of pseudonym. If you have changed it once it may be normal, perhaps you did not like the name.
But if it has been changed four or five times, it is even stranger. If he has also deleted all his previous publications. Or if they have not even been deleted and they used to talk about soccer and now they talk about politics. Then one should be even more suspicious.
Unlike other networks, Twitter has a free open API that allows you to connect and extract information about an account, its followers, tweets, etc.
To access this tool it is necessary to have some computer skills and to be able to handle:
- Phyton or R libraries.
But for those not initiated in programming, there are other tricks that are much simpler and more accessible.
- Using Twitter’s advanced search form it is possible to search by dates, places, hashtags, names, etc.
- Another very interesting way to find out if an account has changed its name is by entering a simple query in the Twitter search engine:
- to:accountname until:End of the year in which it was created in year-month-day format.
Are the tools to detect them useful?
There are many online applications that allow the analysis of accounts. Which have recently been applied to examine the possible fraudulent followers of some politicians.
However, these mechanisms provide less reliable data as the number of followers increases. Since they only take a sample of the latest followers:
From about 300,000 onwards, it produces nonsensical and erroneous biases. Even if they are paid.
If it were a random sample of everyone, it would be worth a little more. But, if an account has 1.3 million users and only the last 2,000 or 3,000 are analyzed, it produces a huge bias.
In the case of Twitter, analysts can get much more accurate data by downloading all the account information through the aforementioned API.
It’s a laborious exercise that can take several hours, but one whose examination proves more accurate.
And what can the user do about fake profiles on social networks?
We asked Facebook, Instagram, Twitter and LinkedIn about these practices. They emphasize that this type of behavior is prohibited or penalized on their platforms.
To combat them, they use human and automatic mechanisms through AI to detect and close these accounts. These protection systems have been increased in recent months to combat a wave of fraudulent activities that have arisen in the wake of the coronavirus crisis.
Hate speech is protected in the US by the constitution. It is their interpretation of freedom of speech, which admits absolutely everything but a direct physical threat.
In Europe, however, they do not allow, for example, attacks on groups that have been historically marginalized. Such as LGBT or xenophobia.
People ask themselves: how is it possible that there are accounts where so many barbarities are said? The question lies in the fact that who decides is an American company, which has its editorial line.
Although it is a content manager, it suits them very well that the law interprets them as a medium. Because they are not responsible for the content.
What can a user do? Cry and pray, because blocking on Twitter is useless. It does not prevent them from continuing to harass, unless there is a rogatory commission from the commission of Spain to a US judge saying that this gentleman is denounced and tell us who he is. Otherwise, you are not going to find out in life.