Turkish Plan to Muzzle Social Media Delayed by Pandemic

As Turkey, like the rest of the world, struggles with the coronavirus pandemic, its government plans to take another step to further restrict digital rights in the country.

A draft law will create new responsibilities for answering the government’s demands on their content for social media giants such as Twitter, Facebook and Instagram and popular messaging apps like WhatsApp and Messenger.

The law on social media was dropped from the parliamentary schedule on Tuesday to make way for more urgent bills on the economy and health amid the COVID-19 pandemic. But civil society groups and opposition parties fear it will be back before long.

Human rights watchdogs, experts and the opposition suspect the government of President Recep Tayyip Erdogan is using the coronavirus crisis to place further controls over social media.

Experts warn that the planned measures would have serious consequences for tech companies’ activities, and may result in some leaving the country.

The draft law on social media has been sent to the business world and unions for consultation, but the opposition is sure it will come back to parliament soon.

“Erdogan’s intention is to close down the social media with this draft law. They will try to bring the draft law [back to parliament] at the first possible chance,” Garo Paylan, an MP from the pro-Kurdish People’s Democratic Party, HDP, told the media on Tuesday.

A day earlier, the director of the Turkish branch of Human Rights Watch, HRW, Emma Sinclair-Webb, wrote: “Not content with simply cracking down on individuals for critical social media posts, Erdogan’s presidency is now intent on using the COVID-19 crisis as a pretext to exert direct control over social media platforms.”

Emre Kursat Kaya, a security analyst with the Istanbul-based Centre for Economics and Foreign Policy Studies, EDAM, says times of pandemics are usually compared to wartime periods for a reason.

“People are pushed to make a choice between their individual freedoms and more security from public authorities. Most of the time, it is the latter that prevails. Not many questions are asked and debates around issues are mostly avoided as the crisis requires rapid responses,” Kaya told BIRN.

New law creates long list of obligations


Turkish President Recep Tayyip Erdogan in Ankara, Turkey, 2020. Photo: EPA-EFE/STR

The draft law obliges foreign social media companies with high internet traffic to appoint an official representative in Turkey to answer authorities’ demands concerning the content on their platforms.

Companies will need to respond to communications from the authorities about their content within 72 hours and compile and notify officials of all removed or blocked content in three-month periods, the draft law says.

More importantly, the companies will also be asked to store data belonging to Turkish users within the country.

If they fail to respond to official requests within 72 hours, they will face penalties up to 135 million euros. Companies that do not compile the removed or blocked content, or do not store data in Turkey, could be fined up to 675 million euros.

The draft law also says that companies that do not follow the government’s new rules could face having their bandwidth halved after 30 days by a court order, and then reduced by 95 per cent if they continue to flout the rules for another 30 days.

“The Turkish authorities have long demanded to have official representatives of online service providers,” Kaya noted.

“This demand was linked to a wish to accelerate the removal of unlawful content from online platforms. But even without having a representative in Turkey, these platforms tend to respond to removal requests quite rapidly, and faster than the 72 hours expected by the text,” he added.

Kaya said the first two aspects of the proposed law would not have such a big impact on how fast content is deleted, but “will only add another layer of pressure on online service providers by taking their representatives as responsible”.

He added: “What’s more worrying …  is the third aspect, which basically requires data localization from online service providers.

“This is highly problematic as there is no precedent of such action from these global companies and this could result in them simply leaving the Turkish market,” he continued.

Companies may quit market rather than obey


Two Turkish women try to get connected to the Twitter in Istanbul, Turkey. Photo: EPA/TOLGA BOZOGLU

Taylan Yıldız, a former Google analyst and member of the Istanbul Municipal Council from the opposition Good Party, said that the draft law has many open-ended articles, and it will mostly affect people with pro-opposition ideas and opposition parties.

In March alone, 433 Turkish citizens were detained because for social media posts that allegedly spread fake or manipulative news on the coronavirus pandemic.

At least four people were arrested or fined for their social media posts, including Fikri Saglar, a former lawmaker from the main opposition Republican People’s Party, CHP.

A legal investigation was also started against Omer Gergerlioglu, an HDP MP, because of his social media posts on the effect of the pandemic on Turkey’s overcrowded prisons.

“Companies are left with only two options. They will stop their operations in Turkey, or they will block every content following the government’s complaints,” Yildiz wrote in his personal blog.

Yildiz said that if social media companies withdraw operations from the country, “Turkey will become introverted and will face a disconnection with the rest of the world”.

Turkey previously blocked several social media companies because of their refusals to delete some content.

Court rulings blocked Twitter several times in 2014, though the ban was later lifted following an agreement between Twitter and the government.

As of 2018, Twitter reported that the Turkish government accounted for more than 52 per cent of all content removal requests worldwide; Twitter only answered 4 per cent of the government’s requests.

Turkey also banned the social information platform Wikipedia for more than two-and-a-half years because of content that the government wanted removed. In a surprise decision, the Constitutional Court lifted the ban on January 15, 2020.

“If the legislation passes as it stands, the main issue for social media providers will be the demand for data localization. This is practically impossible, as it would mean an additional financial burden and unavoidable security risks for these companies,” Kaya, from EDAM, said.

He said the demand for data localization was not unique to Turkey, so if companies concede it to Turkey, they will face pressure to do the same for many other countries. “Sadly, this will probably result in many of them leaving the Turkish market,” he predicted.

The worldwide online payments system PayPal ceased all of its operations in Turkey for similar issues in 2016. “Facebook, Twitter and WhatsApp could follow in PayPal’s footsteps and disable the use their applications in Turkey,” Kaya concluded.

‘Vox Populi’: How Serbian Tabloids and Twitter Bots Joined Forces

Tweets by a more than 8,000-strong ‘troll army’ promoting Serbia’s ruling party and President Aleksandar Vucic regularly found their way into news stories published by Serbian media in the last couple of years before Twitter took them down last month.

Twitter deleted 8,558 accounts engaged in “inauthentic coordinated activity” – some 43 million tweets criticising the Serbian opposition, independent media and individuals critical of Vucic and his Progressive Party rule.

But the bots were not alone.

Analysing just five of the thousands of accounts, BIRN found their tweets were embedded in stories published by the likes of pro-government tabloids InformerKurir and Espreso at least 23 times, suggesting the total number across the network may run into the hundreds.

The tweets were often presented as supporting evidence of the unpopularity of Vucic’s opponents; others were picked up by both Serbian and Russian media as proof of the popularity of Russian President Vladimir Putin at the time of his red-carpet visit to the Serbian capital, Belgrade, in January 2019.

“Weighing in on Twitter disputes and dogpiling onto opposition tweets did not just alter the Twitter landscape in favor of SNS-aligned figures and to the detriment of the opposition,” the Stanford Internet Observatory, a US-based research, teaching and policy program that looks at abuse of information technologies, particularly social media, said in a report in early April.

“In some cases, these tweets would get taken up by web publications as “organic” critical content,” it said, noting that some stories cited tweets from multiple accounts in the network.

In the April 2 report, “Fighting Like a Lion for Serbia”: An Analysis of Government-linked Operations in Serbia, the Observatory said that another important function of the deleted accounts was to “push out links to content on SNS-aligned news websites,” including sns.org.rs and vucic.rs [the official websites of SNS and Vucic, respectively], as well as media outlets such as informer.rs, alo.rs and pink.rs, all staunchly pro-Vucic.

The report, for example, cited a tweet by the editor-in-chief of Informer, Dragan Vucicevic, in which he criticised opposition politician Borko Stefanovic. The tweet was replied to 64 times by the troll accounts.

“This kind of propagation suggests that the network’s influence extended beyond Twitter—although it is impossible to assess the extent of this influence with much precision,” the Observatory wrote.

Snjezana Milivojevic, professor of Public Opinion and Media Studies at Belgrade’s Faculty of Political Sciences, said the Twitter bots and pro-government media were “parts of the same strategy”.

“The Internet is a large free space, so, by directing attention, bots help to prevent the dispersal of the public and help friendly media such as Informer, Pink and Alo to function as a well-run factory of the same fake news,” Milivojevic told BIRN.

Network built to boost retweets and reply counts


Some of the tweets that ended up in mainstream media such as Epreso, Kurir, Informer and other media outlets. “Local government in Cajetina put a mortgage on a parcel where the bones of World War II victims remain. Unbelievable what these people are capable of. Stamatovic, aren’t you ashamed?”Ivan Ilic, wrote on Twitter, later was republished in Informer. Illustration: BIRN

According to the Stanford report, one of the top three bot accounts taken down last month operated under the name ‘Mirjana Kujovic’ [@1kujovic].

The account’s tweets found their way into Serbian and Russian media more than once. Following Putin’s 2019 visit, the Russian website fontanka.ru cited a January 17, 2019 Kujovic tweet as evidence of the warm welcome Putin received.

Months earlier, in October 2018, a negative comment made by the Kujovic account under tweets by Serbian opposition politicians Bojan Pajtic and Vuk Jeremic was then embedded in a story by the tabloid Espreso.

The same month, another tabloid, Srbija Danas, published a Kujovic tweet criticising academic Dusan Teodorovic, a founder of the opposition Movement of Free Citizens, PSG.

Kurir also got involved, quoting another later-deleted bot popular with pro-government tabloids in Serbia – Ivan Ilic [@grofodValjeva]. 

The more than 8,500 accounts deleted by Twitter “worked steadily to legitimate Vucic’s policies and undercut public support for his opponents,” the Standard Internet Observatory wrote.

The accounts tweeted more than 43 million times – 85 per cent retweets.

While some were active in 2009, within months of the Progressive Party’s founding the year before, the network began ramping up its activities in mid-2018, the Stanford report said, right before the start of large, regular anti-government protests under the banner “1 of 5 million’.

The average number of followers attracted by the accounts was just 66, but combined they reached roughly 2.3 million Twitter users. @belilav11 and @1kujovic racked up 12,167 and 10,867 followers and more than 330,000 and 390,000 engagements respectively.

Engagement, however, was not the primary purpose, the report said.

“…they existed primarily to boost retweet and reply counts for other accounts,” it said. “This was consistent with the political aims of this network, which revolved around artificially boosting Vucic and his allies on Twitter.”

The network and its media allies, Milivojevic said, were working to manipulate the Serbian public.

“With 43 million messages [tweets] in which someone is praised or criticised, that manipulation also entails a decline in trust in the media by erasing the boundary between truth and lies,” Milivojevic told BIRN.

And the bots work in concert with genuine, popular Twitter users and pro-government tabloids and broadcasters, she said.

“What is published in tabloids is taken over by influential Twitter users… Then anonymous bots retweet and spread it, and from there on their tweets are going back to informative talk shows, where politicians or analysts bring them in [printed] and show them around,” Milivojevic said, referring to the Pink TV talk show Hit Tvit [Hit Tweet].

‘Like the plague’


Serbian progressive party (SNS) leader Aleksandar Vucic (front- C) addresses the media at a polling station in Belgrade, Serbia. Photo: EPA/ANDREJ CUKIC/ANDREJ CUKIC

Andrej Petrovski, head of tech at the SHARE Foundation, a Belgrade-based digital rights NGO, said Twitter’s deletion of so many accounts should serve as a warning to Serbia’s ruling party.

Creating and managing such a vast network takes a lot of time and people, he said, people he described as members of the SNS “party machinery” each running at least 10 Twitter accounts.

“Twitter made it clear with this move: if you do it all over again, we will do the same, and then you will think whether you want to invest that amount of time, effort and money again knowing it can all disappear overnight,” Petrovski told BIRN.

Some, however, say the bots are back already, promoting the party line amid the COVID-19 pandemic.

“They are like plague now,” said Jovana Gligorijevic, a journalist with the weekly political magazine Vreme and a frequent target of the SNS bots.

“They are all parts of the same machinery,” she told BIRN. “They create a fake vox populi [voice of the people].”

Gligorijevic said that whenever she uses the words ‘Aleksandar Vucic, ‘minister,’ ‘SNS’, ‘the president’ or ‘the prime minister’, she is bombarded by insults and negative comment, to the degree that she once deleted her account. 

“The bots react on those key words,” Gligorijevic said. “This is one network for absolute media control.”

BIRN editor Slobodan Georgiev has also been ensnared.

“First they insult you on Twitter, then that is published in tabloids and then you end up in the ‘analysis’ on tabloid TV stations,” he said.

A bigger problem, however, “is that they make you an ‘enemy’, and that comes directly from the top of SNS, which leads these bot divisions,” he said. And that makes journalism difficult.

“Then you are labelled a danger to the state and people working in the system start avoiding you and stop responding.”

BIRD Community

Are you a professional journalist or a media worker looking for an easily searchable and comprehensive database and interested in safely (re)connecting with more than thousands of colleagues from Southeastern and Central Europe?

We created BIRD Community, a place where you can have it all!

Join Now