![]() |
Social Barrel |
In a blog post last week, Facebook reveal information about how they find terrorist accounts, and in particular about the actual, human staffed counter-terrorism team they have on staff. Said team apparently contains more than 150 people, many of whom come from law enforcement backgrounds, rather than tech. Facebook claim that 30 different languages are spoken within the team, and beyond law enforcement, it also includes former prosecutors and an emergency response team.
Facebook have also pointed towards their partnerships with national governments, law enforcement authorities and, in particular, NGOs. Facebook have worked with NGOs to develop counterspeech, a kind of anti-propaganda that presents a more preventative approach to counter-terrorism - stopping the ideas being spread before any threat of violence has a chance to emerge. Their co-operative summit with Google, Microsoft and Twitter is also mentioned.
Of course, it wouldn't be a Facebook open letter without at least some mention of AI, and sure enough there's a section dedicated to it. In this case, it details the workings of their image matching and language recognition systems, as well as 'clusters' - content which isn't directly connected by any one group or page, but all relates to the same extremist ideology. Facebook also claim that they can detect and remove accounts made by repeat offenders much more quickly than before.
This blog post is a fairly direct response to increasing governmental pressure to take further action. Several national governments have laid out plans to fine Facebook for failing to act against terrorists and others have vowed to place stricter regulations not only on Facebook, but on the internet as a whole. For many, Facebook's counter-terrorist wing won't be enough to satiate them.
Post a comment