On Monday morning I met with a group of activists who live under authoritarian regimes. The delegation had been brought to San Francisco by the nonprofit Human Rights Foundation as part of a fellowship focused on the relationship between activism and Silicon Valley. And the big question they had for me was: why do social networks keep taking down my posts?
The question caught me off guard. For every story in this newsletter about an activist’s post wrongly (and often temporarily) being removed, there are three more about the consequences of a post that was left up: a piece of viral misinformation, a terrorist recruitment video, a financial scam, and so on. As I wrote in 2018, we are well into the “take it down” era of content moderation
Sometimes the activists’ posts came down because their governments demanded it. Other times the posts came down because of over-cautious content moderation. Increasingly, the activists told me, social networks were acting as if they would rather be safe from government intervention than sorry. And whenever their posts and pages came down, they said, they had very little recourse. Facebook does not have a customer support hotline, much less a judicial branch. (Yet. More on that below.)
The activists’ concerns were fresh in my mind when I read about the weekend’s removal of Instagram accounts in Iran that expressed support for the Iranian general Qassem Soleiman, who was killed by the United States last week. Like a strong antibiotic, it appears that Instagram’s enforcement action wiped out both accounts tied to the ruling regime and the posts of everyday Iranians.
As part of its compliance with US law, the Facebook spokesperson said the company removes accounts run by or on behalf of sanctioned people and organizations.
It also removes posts that commend the actions of sanctioned parties and individuals and seek to help further their actions, the spokesperson said, adding that Facebook has an appeals process if users feel their posts were removed in error.
GoFundMe also removed at least two fundraising campaigns for passengers on the Ukrainian flight brought down by Iranian missiles, only to later reinstate them, my colleague Colin Lecher reported at The Verge
. But Twitter, on the other hand, said it would leave posts up so long as they complied with the company’s rules.
The confusion is to be expected. Legal experts disagree on the extent to which sanctions require tech platforms to remove user posts, and the issue of Iran in particular has been giving companies fits for years. Here’s Lecher in The Verge:
While recent news has put the focus on Iran, it’s hardly the first time tech companies have mounted a zealous response to sanctions. Last year, GitHub restricted users
in several countries under US sanctions.
Iran, which has faced sanctions for years, has regularly had tech companies limit use in the country in response to US policy. In 2018, Slack deactivated accounts around the world
that were tied to Iran, in a move that stretched well beyond the borders of the country. Apple took several popular Iranian apps off its store in 2017 in the face of US sanctions
. At the time, Apple issued a statement that’s still relevant: “This area of law is complex and constantly changing.”
At the same time, once again people around the world are waking up to the reality that their speech is governed by actors who are not accountable to them. Instagram has users but not citizens. Executives in California will decide what can be said in Tehran.
Of course, there’s vastly more free speech on Instagram than in a country like Iran, where speech is brutally repressed. But as the activists shared with me on Monday, the ramifications of social networks acting as quasi-states to reshape political speech in their countries are significant. And their struggles to appeal unjust content removals are real.
The good news is that later this year, Facebook will launch its independent Oversight Board
: a Supreme Court for content moderation that will allow users to appeal in cases like the activists’ and the Iranian citizens’. One of the board’s rules will be that cases selected for review will include at least one person from the region in which the case originated. That’s not quite a democratically elected representative — but hopefully it bolsters the board’s accountability to Facebook’s user base.
There are still many questions about how the board will work in practice, and whether it can serve as a model for quasi-judicial systems at other companies. But hearing the activists’ stories today, and reading about the confusion over sanctions in Iran, it seemed to me that the board can’t launch quickly enough.