Maxim Behar for Eurocom TV: "Only artificial intelligence can defeat trolls on social media"

Maxim Behar for Eurocom TV: "Only artificial intelligence can defeat trolls on social media"


Maxim Behar, Founder and CEO of the leading PR company M3 Communications Group, Inc., spoke to Eurocom TV with Slav Mehandzhiyski on trolling, bots and how to recognize fake profiles and their propaganda. Professor Maria Neykova, Kristina Hristova, chairwoman of the Coalition for Media Literacy and MEP Andrei Slabakov also had their say.

Interviewer: War and battles between trolls on social media. This is the pandemic that is flooding the digital space and there’s no vaccine for it. As the elections and the beginning of the elections campaign approach, we are witnessing more and more frequent publications on the network, but also more and more sharpening opposition and attempts to impose an opinion.

Due to the pandemic, more and more campaigns are choosing the digital space, where it is also essential that the rules of the game apply for everyone. However, the people also known as trolls on the internet, do not play by the rules - they are difficult to trace, merging among all users, but they manage to have a strong impact, which is why they are often used.

What do we need to know about them and what are the consequences?

Maxim Behar: Trolls who spread fake news or comment on someone else's identity react mostly in the area of hatred and hatred. Since 2004, when Facebook was launched on May 4, they were born and it was very easy to predict that there would be so many people who would abuse social media, and they will be growing. Facebook currently has 2.8 billion users, and it's normal for some of those nearly 3 billion people who have social media in their hands to want to abuse it.

Since you mentioned this social channel, Facebook is undeniably the most recognisable, transparent and personalised media possible, and it’s quite easy to recognise when someone is trolling. Trollers don’t have many followers, but they use someone else’s identity, including images and name in attempt to provoke a reaction. In addition to that, the second you report the troll or the fake profile, their profile is gone within 5 minutes. There are more social media channels like Instagram, Twitter, TikTok and the list goes on, where trolls are hard to be identified.

Interviewer: Is there any method to identify them?

Christina Hristova: Yes, there is. In most cases, these profiles do not have a real person’s photo. They easily be found via Google reverse image search. Most often, they steal photos from advertising bases and it crystal clear that these are not real people, because the photos are too good to be true or do not have a profile picture of a person at all.
Lately, we’ve been noticing a tendency, in which trolls use the Bulgarian flag, a horse or a landscape as a profile picture. It’s important to notice that these fake profiles do not have personalised photos. The whole idea behind the trolling refers to a certain cause or idea. For instance, the theory that COVID 19 is spread through 5G network and all the activity is around that topic or beautiful sunsets and animals, but there are no real human experiences on the wall, and they do not have too many friends.

Interviewer: We used expert’s knowledge to easily recognise fake profiles. In the comment section, we can always find potential trolls, but what draws the attention is that their comments provoke the most reactions amongst consumers. We did a profile check with a place of residence Jordan, but when we refer back to Google reverse search image to find out if this was a real profile, we found that the living in Jordan is famous model from Greece. The next profile that rang some bells was used mainly for games, and later the topic became political. It was more than clear that the names between the account and the link above also differ. The main question against the background of the upcoming elections is how widespread such profiles are in Bulgaria.

Professor Maria Neykova: In relations to the parliamentary elections, it seems to me that two things are particularly important. The first is the loss of common ground for discussion and debate. This common space is spreading and can serve as a trolling place. The second is the technological change, which is much greater than the one that has taken place in recent years. I was impressed by the fact that one of the party formations has already released a video claiming that in the long-running election campaign there is a similar interference with the creation of fake Facebook profiles, in which the known as the "props" are circulated. The servers are abroad as the party formation claims. I cannot see how this is can be overcome.

Kristina Hristova: What we are noticing is that the main political parties use trolls and they are largely on an equal footing. I do not know to what extent they can really influence, as the parties with the highest turnout have an equally strong presence among troll groups.

Maxim Behar: Of course, none of the parties will admit that there are trolls, but it is a public secret that each party has some people who stand in their offices, follow instructions and discredit the opposing party or the competing political force, and at the same time they write positive things for the political force that pays them. Most people understand this. About 30 percent of society can find their way around very easily. However, the remaining 70 percent read the opinions more than the information on the post itself, on any social media. Indeed, this is the purpose of trolls - to fill, to flood the space around a news or a post, so that the majority of readers focus on the opinions of these trolls, rather than the news itself.

Interviewer: Two months prior the elections, 4 April, “Yes Bulgaria” announced that there were number of trolling comments under its post. The party linked the profiles with Pavel Valneve, who is deputy chairman of the party “Republicans for Bulgaria.” However, in many cases it remains difficult to whether this it’s a fake profile and real person. Political divisions in society make it difficult to determine whether or not sharp comments are trolling. The biggest saga around the trolls became the US elections in 2016. The biggest saga around the trolls became the US elections in 2016. In the first place, the city of Veles in our South-western neighbour, Macedonia became famous for spreading fake news to US voters, and in 2020 website became more active for the new US elections. In Veles, populated with around 40,000 people At least 140 websites with names containing Donald Trump were registered. The sites widely disseminated those of the Republican candidate, filling social networks and groups with their content.

This avalanche of fake news, which has spread across the globe thanks to social media, has provided Trump with extra voters, and funding for trolls.

Mark Zuckerberg (2018 video): One of my greatest regrets in running the company is that we were slow in identifying the Russian information operations in 2016. We expected them to do a number of more traditional cyber attacks, which we did identify and notify the campaigns and they were trying to hack into them. But we were slow to identify the type of new information operations. We were able to remove tens of thousands of accounts as a precaution before they did any harm. There are people in Russia whose main task is to try exploit our system and they will try to get better, which means that we also have to get better.

Maxim Behar: You remember the very recent mention of Donald Trump being blocked. I think it was a big mistake and they themselves, as well as Twitter, realized that it was a big mistake, but it was still necessary at the time. Facebook is very easy to recognise by keywords and has very strong software that deals with it. Artificial intelligence, which recognizes keywords, sends alerts, bans the account for a while. And when you don't understand and keep insulting, using swear words or any other words, they block you.

In my opinion, artificial intelligence is one of the tools we can use to fight against fake news, hatred, harsh language or pedophilia, which can be found in the criminal code as criminal acts anyway.

Professor Maria Neykova: We have a set of very complex questions to solve: What is freedom of speech? When will these technology media giants - Twitter, Facebook - be regulated? In fact, the word you used "turbulent" is in the same position as the media, because we still cannot fully comprehend and assess the impact of new technologies, including social media.
Interviewer: The European Union has started introducing regulations as the fight against fake news is one of the main priorities of the Union.

Ursula Von Der Layen (Video 2021): Activists warned of the business models of large technology companies and the implications for democracy. They described how the economic laws of social media are eating away the foundations of our society, about how fake news spreads 6 times faster than real news. They rely on Europe because the Union has proven that it can set standards that are respected in color, such as the "Data Protection Rules".

Professor Maria Neykova: In the European Union, on the recommendation of the European Commission, the broader term "misinformation" is used to mean information that can be shown to be false or misleading and was created and disseminated for economic gain or to deliberately mislead the public. the public. In addition, we stopped using fake news because of politicians who, if they didn't like something, tended to dismiss it as fake news. A number of studies, including in Bulgaria, have found in recent years that politicians are one of the big generators of fake news. In this sense, with a disintegrated public field for discussion, I believe that everyone will trust the party or the representatives of the party they have decided to vote for.

Christina Hristova: Because a truly objective news, a message built on real facts and a constructive policy has no chance against hysterical shock conspiracies that carry many threats. As you know years ago, thanks to this false news, a very large number of people stopped getting vaccinated, young people stopped vaccinating their children. So there is even a real threat to our health from fake news.

Maxim Behar: If you prevent profiles from being created without the necessary information from the creator, this will drastically reduce fake news and what is known as "trolling" or “trolls” who put a photo from somewhere, add a few artificial friends and photos. This is very common. In America, England, France, Germany, people are much more cautious. They look much more closely who is behind the profile and what photos they have. In Bulgaria you can catch a picture of a half-naked girl or very authoritative and intelligent person. There was a case of a man who had recently been released from prison and had been selling fake plane tickets. He had raised a lot of money from people through his fake name and photos. They saw him on Facebook and immediately believed him. We need to be much more careful and this will happen over the years.

Interview: Before the introduction of resolution of the European Parliament, MEP Andrey Slabakov shared about the problems with fake news, which he personally experienced.

Andrey Slabakov: If we have to talk about freedom of speech, I think the main problem is not freedom of speech. The main problem at the moment has become misinformation and fake news, especially on the Internet. Misinformation platforms are everywhere. I can tell you that I have not given interviews since May, and in the last few months I have read dozens of my interviews given by me, as well as statements that some of the information sites allow themselves to sign on my behalf.

Interviewer: The relocation of social media campaigns raises questions about whether trolls and fake news can influence public opinion and if they can be regulated.

Mark Zuckerberg (video 2018): For the language of hatred, I feel optimistic that for a period of 5-10 years we will have artificial intelligence that will be able to cope with the linguistic nuances of different content and will be more accurate with the reports, which are sent to our systems. We have policies that prevent subjectivity, but until we introduce artificial intelligence, mistakes are more common than what I would ideally like.

Professor Mariya Neykova: People deepen their beliefs over the years, and eventually they become their characteristics. If a person is a clear supporter of a party, they will not change their beliefs. Rather, it is aimed at the hesitant. Furthermore, we must not forget that the situation is such that suspicions of low voter turnout are very real. We don't really know what will happen on April 4th.

Maxim Behar: It’s a simple recipe. In my opinion, fake news should be criminalized in law. If you push a person on the street or enter a house with a burglary, all these are criminal acts are subject to appropriate treatment by the penal code. However, if you insult someone or pour a bucket of lies against them, if you break their business and a bunch of other consequences that the writings of a fake profile that creates fake news can have, all this goes unpunished. Why? And I am sure that sooner or later false news, untruths, rude language and hatred will be included in the penal code as criminal acts. This will drastically reduce their spread.

Watch the whole video here.