To suppress coronavirus misinformation, Twitter began more strictly applying its rules to heads-of-state during the pandemic. Platforms’ aggressive efforts to counter misinformation reset expectations for how they should treat government leaders. In January, Twitter suspended then-President Donald Trump for allegedly inciting violence at the Capitol. Demonstrating that the new rules apply universally, Twitter recently removed a threatening tweet by Nigerian President Muhammadu Buhari. Nigeria responded by blocking access to Twitter and demanding it create a local subsidiary, an approach to social media regulation seemingly borrowed from Turkey. Widespread adoption of the Turkish model will imperil internet freedom everywhere.

In early June, Nigerian President Muhammadu Buhari used Twitter to threaten Igbo separatists, referencing his service in the Biafran war.

“Many of those misbehaving today are too young to be aware of the destruction and loss of lives that occurred during the Nigerian Civil War. Those of us in the fields for 30 months, who went through the war, will treat them in the language they will understand.”

Shortly afterwards, Twitter deleted the post for violating its policies against abusive and threatening language. Seemingly in retaliation, the Nigerian Federal Ministry of Information and Culture announced — via Tweet—that it had “suspended, indefinitely, the operations of the microblogging in social network service, Twitter, in Nigeria.”

In subsequent press conferences, Nigeria’s Minister of Information and Culture, Lai Mohammed expressed skepticism of Twitter’s motives, “Twitter’s mission in Nigeria is very suspect, and they have an agenda.” This was perhaps a reference to Twitter’s support of the Nigerian anti-police brutality movement #EndSARS in the fall of 2020, another perceived violation of national sovereignty by the corporation. Governments accustomed to deference from commercial firms have looked askance at corporate support for local social justice movements.

Days later, the minister announced that, in order to operate in Nigeria, social media platforms “must first be a Nigerian company” and additionally “be licensed by the broadcasting commission.” In order to resume operations in Nigeria, Twitter, and other social media firms, would have to create local subsidiaries subject to Nigerian law.

Over the past decade, Turkey has repeatedly restricted access to social media services in order to extract concessions from them. It eventually required firms to employ local representatives beholden to Turkish law. Nigeria’s recent approach to social media companies resembles the progression of Turkey’s heavy-handed style, compressed into a period of two weeks: a countrywide service ban which is relaxed in exchange for particular concessions from platforms.

Turkey’s distinct approach to internet censorship began in 2007, when government officials discovered YouTube videos mocking Mustafa Kemal Ataturk, the founder of modern Turkey. The videos featured images of Ataturk’s face pasted onto a monkey, Ataturk depicted as a clown, and claims that “Ataturk is gay.” The Turkish government viewed these videos not as lighthearted comedy, but a serious violation of its laws against “insulting Turkishness.” In response, an Istanbul court issued an order banning YouTube in its entirety. State-run internet provider Turk Telekom readily complied. A few hours after being blocked, YouTube removed the offending videos, prompting a Turkish newspaper to proclaim “YouTube got the message.” YouTube’s concession prompted the Istanbul court to rescinded its order, but other offensive videos were brought before other Turkish courts, which blocked YouTube seven more times in 2007 alone.

Turkish authorities applied what had worked with YouTube to other platforms. In 2015, videos of the kidnapping of Turkish government official, Selim Kiraz, were widely shared on Twitter, Facebook, and Periscope. The Turkish government responded by banning the offending websites for violating their Internet Act’s rules against “anti-government propaganda.” After cutting access to the platforms altogether— costing them millions in lost ad revenue—Turkey’s Information and Communication Technologies Authority requested the removal of the videos. When the platforms complied, access to their sites was restored. By bluntly precommitting to punishing platforms for failing to toe the line, Turkey was able to informally shape their content moderation policies.

Turkey also made use of platforms’ “country-withheld content” tools to limit the speech available to Turkish internet users. Although these tools were originally developed to comply with German bans on fascist symbolism, Turkish authorities have used them to suppress domestic criticism. The expanding use of this capability illustrates how compliance with one illiberal request paves the way for further restrictions. In 2018, Turkey requested more than 5000 such removals from Twitter alone.

Turkey formalized its censorious arrangements in January 2021 with a new social media law “requiring sites with more than a million daily users to appoint a local representative by next month to enforce court orders to remove content.” The legislation requires platforms to select employees who are subject to Turkish law as local representatives. These employees are legally responsible for ensuring platforms’ compliance with Turkish court orders. In effect, the law requires the designation of a hostage, or at least a scapegoat. Compared to service shutdowns, this mechanism offers Turkey a much more subtle instrument for shaping platform policy, though it builds on the success of earlier blunt methods. Turkey’s YouTube bans had already demonstrated that it was willing to go without the platforms entirely. Most platforms readily complied with the new law, but Twitter, Periscope, and Pinterest appointed representatives only after being punished with an advertising ban.

The Turkish government’s willingness to suspend market access gives its demands their force. There’s only one Turkish government, the gatekeeper to the Turkish market, while there are many social media platforms that want to do business there. If any one platform fails to comply with Turkey’s demands, it will lose market share to its competitors. Turkish citizens, on the other hand, have little reason to protest the loss of a single social media service. Unless platforms are willing to cede the Turkish market to their rivals, or find technical means of resisting service shutdowns, they will find it structurally difficult to resist these tactics.

It is therefore unsurprising that other illiberal governments are adopting the Turkish approach to internet governance. These governments want both the economic benefits of the internet and the speech control they enjoyed in the pre-digital era of state-run television. However, platforms have no responsibility to offer the neutered versions of their services demanded by such governments. Individually, social media platforms have little reason to privilege the desires of governments over those of their users, particularly when doing so degrades the quality of their products and compromises their values. Most Nigerians want an uncensored internet. Platforms are in the business of providing “voice,” while these governments assert a fundamentally incompatible right to control political expression.

Whatever their private rules, social media platforms have long exported American speech norms, allowing the citizens of less liberal states to avoid local censors. For better or worse, platform moderators neither knew nor cared about local sensitivities. However, as the internet has become more important within less free societies, platforms face increasingly sophisticated local censorship demands.

Nigeria’s recent actions bear striking similarities to Turkey’s method of internet censorship. Whether learned or hastily borrowed, the Nigerian proposal illustrates the danger of platform’s acquiescence to government ultimatums. Paying one ransom simply invites more demands.