A report from the Christchurch Call, where the future of "anti-extremist" moderation was debated at the highest levels

This week's Christchurch Call event in Paris brought together politicians, tech execs and civil society to discuss means of "countering violent extremism" online; it was convened by New Zealand Prime Minister Jacinda Ardern in the wake of the deadly white supremacist terror killings in Christchurch last March.

My Electronic Frontier Foundation colleague Jillian C York was there as part of the civil society delegation, and her report on the event highlights some of the genuinely positive outcomes from the event (a commitment to "strengthening the resilience and inclusiveness of our societies" and a mandate for tech companies to be transparent in their content moderation); as well as some of the not-so-good conclusions (a lack of distinction drawn between services like Facebook and infrastructure like DNS when it comes to conscripting companies to reduce violent extremism).

Most disturbing, though, was a commitment to requiring algorithmic filters of human expression, something hinted at in the terrible, hastily enacted Australian bill passed in response to the terrorist attack. The use of filters to curb bad speech has gained widespread acceptance in policy circles in the past 12 months, despite the near-total consensus among technologists and computer scientists that this will not work and will have ugly consequences for both human speech and competition.

* The Call asks companies to take “transparent, specific measures” to prevent the upload of terrorist and violent extremist content and prevent its dissemination “in a manner consistent with human rights and fundamental freedoms.” But as numerous civil society organizations pointed out in the May 14 meeting, upload filters are inherently inconsistent with fundamental freedoms. Moreover, driving content underground may do little to prevent attacks and can even impede efforts to do so by making the perpetrators more difficult to identify.

* We also have grave concerns about how “terrorism” and “violent extremism” are defined, by whom. Companies regularly use blunt measures to determine what constitutes terrorism, while a variety of governments—including Call signatories Jordan and Spain—have used anti-terror measures to silence speech.

The Christchurch Call: The Good, the Not-So-Good, and the Ugly [Jillian C York/EFF Deeplinks]