VOX-Pol convened an invitation-only workshop titled ‘Countering Violent Extremism Online and the Impact on Civil Liberties’ Berkman Klein Centre for Internet and Society at Harvard University on 23 – 24 October 2017.
The purpose of the workshop was to explore the challenges and opportunities facing actors engaged in countering violent extremism online, particularly the impact of content regulation on civil liberties.
There were 35 attendees from a variety of backgrounds, including representatives from UNESCO, UN-CTED, OSCE, Facebook, Microsoft, Cloudflare, Committee to Protect Journalists, Institute for Strategic Dialogue, Netzpolitik, Global Network Initiative, Global Voices, Moonshot CVE, and a number of universities, including Columbia University, Harvard Law School, Pennsylvania State University, Dublin City University, and NUI Galway.
The many challenges and obstacles facing CVE online were highlighted during the two-day workshop. These include definitional problems, how to counter violent extremism without undermining human rights, legal and jurisdictional concerns, the pressures on intermediaries to take down content, issues with automated content flagging and removal, lack of transparency and accountability, and the fragmentation and inconsistency of online CVE efforts.
Emerging opportunities for CVE were identified and discussed as well. These topics included the benefits of a multi-stakeholder approach to online CVE, the potential in automated content regulation, building consensus against mandates for government censorship, opportunities presented by insights from ‘big data’ analysis, the need to enhance transparency around government-mandated content take down, online counter-speech, leveraging of ever emerging technologies, and the need for further cross-disciplinary work.
This workshop was connected with a VOX-Pol study being conducted by Central European University’s Center for Media, Data and Society on the same topic.
For a further flavor of the workshop discussions, see the Twitter hashtag #VOXPolCVE.