For a decade, the top search result for “EU referendum” on Google was the political blog EU Referendum. Then it was abruptly displaced by solidly pro-EU media outlets. It appeared that someone at Google had decided that search traffic should be driven to pro-EU sites. Ingrid Carlqvist, a Swedish columnist who covers, among other things, migrant violence, at Gatestone, had her Facebook account deleted after posting a video detailing migrant rapes in Sweden.

These seemingly isolated incidents fit into a larger pattern as Facebook, Google, Microsoft and Twitter helped create and signed a “code of conduct” banning hate speech. Facebook had already become notorious for its political agenda while Twitter had created a Trust and Safety Council filled with extremist left-wing groups like Feminist Frequency to censor the politically incorrect.

Google has historically been a pro-free speech outlier. Its politics have never been ambiguous, but it has eschewed the overt censorship of some of its new partners working to keep the EU free of political dissent. But the code of conduct goes well beyond censorship. The companies will be working to strengthen their “ongoing partnerships with civil society organisations who will help flag content that promotes incitement to violence and hateful conduct”. That amounts to empowering left-wing advocacy groups to dictate content removal to major companies. It means that not only Twitter, but Facebook, Google and Microsoft will get their own Trust and Safety Council. It may be called something else. It may not even have a name. But it will have power. That’s what this really means.

And it’s only a starting point in a larger propaganda initiative.

“The IT Companies and the European Commission, recognising the value of independent counter speech against hateful rhetoric and prejudice, aim to continue their work in identifying and promoting independent counter-narratives,” the press release reads.

Even more than the censorship, the counter-narratives push represents a troubling development.

Left-wing groups won’t just be embedded as censors, but major tech firms will be expected to promote their agendas. And the biggest resource that companies with massive social media platforms have at their disposal isn’t mere money. It’s the ability to decide what is trending and what isn’t.

If a story about Islamic terrorism trends, will Facebook or Twitter be expected to promote a counter-narrative from an Islamic group? How exactly is this any different than traditional propaganda?

“The IT Companies to intensify their work with CSOs to deliver best practice training on countering hateful rhetoric and prejudice and increase the scale of their proactive outreach to CSOs to help them deliver effective counter speech campaigns. The European Commission, in cooperation with Member States, to contribute to this endeavour by taking steps to map CSOs' specific needs and demands in this respect,” the release tells us.

CSO stands for Civil Society Organization. It’s used more often now that NGO carries with it an air of contempt. That last sentence informs us that the CSOs will have “demands.” The European Commission will help leverage and assemble these demands. Meanwhile major tech firms will be working to aid these CSOs in pushing their agenda.

What will this look like? We got a preview of it with Facebook’s "Initiative for Civil Courage Online". Facebook had been facing pressure from Germany’s Merkel who was worried over public outrage at crimes committed by her Muslim migrant arrivals. Censorship was obviously the order of the day.

The Initiative promoted Klick It Out which, in properly Orwellian fashion, urged people to “See It. Report It.” The “It” being “Social Media Discrimination.” And then users were expected to “Klick It Out”. It was a failure. But Facebook and friends are doubling down.

Tech companies love the idea of creating “counter-narratives” because it’s cheaper to throw some money at an NGO or CSO, or to boost their profile, than to invest still more money in censorship. It’s not because they have a bias for free speech, but because active censorship, even when outsourced to poorer countries, which it often is, demands more resources. Pushing an agenda is cheap.

And the goal of companies like Facebook is to increase usage, rather than reduce it, which is why COO Sheryl Sandberg championed “like attacks” in which users flood the pages of bigots with their own speech. But the code of conduct is a thorough rejection of any of that self-interested libertarianism. Censorship is packaged together with agenda pushing. There might be like attacks, but what the EU really wanted was deletion and promotion for the groups that its leaders support. And they got it.

Some fraction of these efforts may be directed at ISIS supporters, but there is no particular reason to be optimistic about that. By putting CSOs first, the message is that this isn’t about counter-terrorism, but about promoting one set of political agendas at the expense of another. Much like Twitter’s Trust and Safety Council, this is about selecting who should speak and who should be silenced.

Programs like these operate under the umbrella of fighting extremism. And extremism, unlike blowing up buses or beheading hostages, is in the eye of the beholder. And the beholder is a tech company standing on the left while looking to the right. The obsession with radicalization treats lawful speech as the precursor to violence. It also assumes that Muslim terrorism emerges from a cycle of extremism between Muslims and critics of Islam. Silence the critics and you stop the terrorism.

European governments, like those of Angela Merkel, are far less worried about Salafists than they are about domestic political dissent. When Merkel berated Facebook CEO Mark Zuckerberg over insufficient censorship, it wasn’t because she objected to the pipeline that feeds Muslims from Germany into ISIS. Muslim terrorism is inconvenient, but political dissent is politically explosive. Social media comprise an alternative organizing force that counters the dominance of media narratives. That makes it a threat.

Attempts to silence more prominent voices like Richard North and Ingrid Carlqvist have run into a backlash, but it’s impossible to rally in support of each ordinary person who has their account shut down or their blog pushed down in the rankings for having politically incorrect views. Social media at their best bring people together. This initiative is about disrupting social organizations that are disapproved of.

It is about preserving the dominance of a government-media narrative while promoting astroturf organizations that try to appear independent, but really echo that very same narrative.

Private companies have the right to determine what content appears on their platforms. But Facebook, Google, Microsoft and Twitter have become part of an alliance with governments and advocacy groups to maintain a particular narrative. They will not simply be removing hateful content. Instead they have undertaken to play a role in putting forward a particular set of ideas by particular governments.

That’s propaganda and it is the opposite of how the internet was meant to be used.

 The deal puts a series of private organizations, backed by EU government power, in charge of determining the content of social media, both positive and negative. Social media were meant to be centered around the user. Instead this deal displaces the user and replaces him or her with the EU.