Facebook tackling terrorism with AI

Web Log: Firm using image-matching tech to weed out Isis propaganda videos

Facebook is using artificial intelligence to stomp out terrorist content

Facebook is using artificial intelligence to stomp out terrorist content

a
 

Aside from its recent industry partnership with Twitter, Microsoft and YouTube to share “hashes” or unique digital fingerprints of terrorism-related videos and images, Facebook has reiterated its efforts to detect and remove such content from the social networking platform through various AI-driven techniques.

Currently, image-matching tech means that when terrorist content eg an Isis propaganda video is detected once, all future attempts to upload and share will be prevented.

Machine-learning algorithms

In addition to this, machine-learning algorithms are spotting language patterns in text that may indicate support for a terrorist group or ideology while other algorithms are deployed to find and remove repeat offenders attempting to set up new Facebook pages or groups.

Facebook’s ownership of Instagram and WhatsApp is also coming in useful, say Monika Bickert, director of global policy management, and Brian Fishman, counterterrorism policy manager, as they describe cross-platform collaboration and data sharing across these platforms to help “take action against terrorist accounts”.

https://newsroom.fb.com/news/2017/06/how-we-counter-terrorism/

a