Turkey’s new emergency law seeks to force websites to reveal user identities

Turkish President Recep Tayyip Erdoğan’s government has proposed a draft law which seeks to attach a series of online censorship measures to an economic aid package aimed at stabilising an economy hit by the coronavirus crisis.

The new law defines ‘social media platforms’ very widely, as “the people or legal entities who allow users to create, view or share data like text, images, voice, location online with the purpose of social interaction,” and states that they will be held responsible for any “inappropriate content” that their users post on their platforms. 

The law will apply to any platform with more than 1 million users in Turkey. The draft law states:

“The abroad-based social network provider that has access to more than 1 million people in Turkey is responsible for assigning at least one authorised person as a representative in Turkey to register the notifications, declarations or requests sent by institutions, associations, legal and administrative offices and also to be responsible for sharing the identity and communication information of this person [who has posted inappropriate content] with the institution. The social network provider openly displays the communication information of this person on their website in a directly accessible way.”

It is very interesting that the Turkish government is aiming to effectively end anonymity on social media platforms. This is very similar to what Western governments have attempted and failed to do already, because anonymity has always been a core part of the internet, and it is unrealistic to expect all social media sites to implement systems to confirm the ID of their users. 

Creating a database that ties social media accounts to official IDs would be very beneficial to authoritarian leaders like Erdoğan, but would also be vulnerable to hacking by criminal organisations. The legislation seems to ask all social media sites to ID their users as if they are entering a nightclub, which is logistically and ethically impossible.

In January I wrote for Byline Times about the British government’s ongoing attempts to end anonymity for UK internet users. The proposed legislation in Turkey similarly seeks to get websites to report on their users, helping the government to police speech on their platforms. Up to now, the Turkish government has been able to force Internet Service Providers to block access to certain websites by legally forcing them to do so. But for sites like Twitter, Turkey needs to file a report which explains how a Twitter account breaks their domestic law. Twitter’s Country Withheld Content policy governs how this request is made.

Twitter publishes transparency reports detailing how many requests countries make to withhold content from IP addresses from their countries. Turkey is the country that issued the most requests in 2019, with a total of 6,073, of which just 5 percent were granted. It seems that the new law is in part a response to the Turkish government’s inability to censor content on websites that are not headquartered in Turkey, meaning they cannot bring domestic legal pressure to bear on forcing censorship compliance. Ahval’s website, for example, is blocked in Turkey, while its Twitter accounts are not, as the Turkish government cannot prove to Twitter that they are breaking any laws.

Article 56 (5) of Turkey’s new draft law states: “Domestic or foreign social network providers that have more than 1 million daily users in Turkey are obliged to keep their data in Turkey.” 

This part of the law seems intended to make it easier for the Turkish government to access data about social media users based in Turkey. Presumably, this would make it easier for them to obtain data on anonymous users of social media who are heavily critical of the Turkish government. Turkey’s Interior Ministry reported that 2,000 social media users had been identified and arrested for “provocative” social media posts related to the coronavirus outbreak at the end of March. This crackdown included people that the Turkish government have accused of owning well known anonymous Twitter accounts, such as the Ankara Kuşu account.

The law also seeks to impose fines on ‘social media providers’ who do not respond to takedown requests. Such fines can be from as little as 100 Turkish lira ($15) to as much as 5 million lira ($746,500). Article 56 (7) states:

“In the event that the content determined by the judge or court decision to be unlawful is reported to the social network provider, the social network provider who does not remove the content within 24 hours despite the notification or does not prevent access to the content is responsible for the compensation of the damages arising.” 

Article 56 (2) also states that the state or institution can apply to criminal courts to limit the bandwidth of websites which fail to comply with orders by up to 50 percent, and then by up to 95 percent after another 30 days. This represents a threat to net neutrality much greater than what has been seen in the United States recently.

One of the problems with this law will be how the Turkish government is going to force foreign social media companies to set up legally responsible offices in Turkey which they are already threatening with substantial fines. In 2016, the Turkish government asked PayPal to move their server operations to Turkey, but instead of complying, PayPal simply abandoned the market.

Another issue is what counts as a ‘social media platform’. Turkey concluded its dispute with the Wikimedia Foundation (WMF), the charity responsible for promoting Wikipedia, in January 2020. Wikipedia was unblocked, and Turkey looked as if it had given up its attempts to force Wikipedia to self-censor. Does Wikipedia fall under this new definition of a social media platform? I consulted two Turkish lawyers on whether the law would apply to Wikipedia, and both thought that the main target of the legislation was big U.S. companies like Twitter and Facebook, rather than Wikipedia.

I used to work for the British subsidiary charity of the WMF, so I followed the Wikipedia block dispute closely. Wikipedia carries no advertising, so they didn’t lose money from the block, and refused to bow to pressure to censor certain pages for Turkish IP addresses. Turkey eventually backed down. Turkey had repeatedly asked Wikimedia to open an office in Turkey and abide by court decisions, but this is not the way that Wikipedia operates. Local groups of Wikipedia editors request recognition by the Wikimedia Foundation in San Francisco and then request funding. The WMF doesn’t open an office itself. It’s more like a franchise. So there was no way that Wikimedia could meet this demand.

The question now is whether Twitter, YouTube, and Facebook will comply with a law which would put them in a position where they are open to large fines and obliged to inform on their users, or whether they will push back against the legislation. Those sites require a lot more bandwidth than Wikipedia does, so they are more vulnerable to financial and access pressure. But as we have seen in Britain with legislation designed to increase state control of the internet, passing a law doesn’t necessarily mean that the law will have any effect.

The opinions expressed in this column are those of the author and do not necessarily reflect those of Ahval.