Crisp Announces U.S. Launch of AI Technology Used to Detect Terrorist Propaganda and Illegal Content on Leading Social Networks

Hundreds of new items of extremist content being created daily online

New York, NY – October 19, 2017Crisp, known for its work in detecting and preventing online sexual predators while helping thousands of brands across the globe manage social media risks, announced today the U.S. launch of an advanced Artificial Intelligence technology used to identify online terrorist and extremist content. The technology – called Capture – is now being used by global social networks to detect new terrorism-related content created online. It has been locating nearly 200 pieces of new, unique pieces of terrorist content every day while operating in beta.

The amount of terrorist propaganda and extremist content being posted online, especially on social media platforms, is a growing global issue. Each individual piece of content has the potential to be uploaded and shared thousands of times, making it essential that it is found and addressed quickly.

Capture works very differently compared to traditional “seek and destroy” moderation methods used by social networks to date. Due to the vast amount of content uploaded every minute, trying to find terrorist propaganda amidst all legal content is similar to finding a needle in a haystack. Instead, Capture works by graphing the social web, finding the places that have been identified by the AI as risk hubs for the discussion and sharing of illegal content, such as dark web forums, discussion boards and messaging apps. It then follows those discussions to locate the source so it can be reported and removed within minutes of being discovered.

“Being able to detect this new content quickly is essential,” said Adam Hildreth, CEO of Crisp. “We are finding hundreds of newly created items every day, but the impact is much wider than that. We know that new terrorism-related content can have a starburst effect, where it is rapidly shared and viewed thousands of times. The Artificial Intelligence behind Capture identifies the highest risk content that is most shareable within minutes, reporting it directly to the appropriate teams at the social networks to action or even automate complete take down.”

The illegal content being discovered by Crisp’s Capture technology can be broken down into the following categories:

  • 73% - non-distressing terrorist propaganda: images and videos which glorify or promote the terrorist cause, but without the use of extremely distressing material 
  • 15% - violent terrorist propaganda: beheadings and bloody gore resulting from terrorist activity shown in the context of glorifying or promoting the terrorist cause 
  • 8% - contains extremist preaching: videos and written pieces that promote hatred or violence against others 
  • 2% - instructional videos on terrorist devices: detailed instructions on how to create explosive, chemical devices or weaponry for use in terrorist acts
  • 2% - incitements to commit terrorist acts: content that directly incites the audience to carry out terrorist acts that would result in the harm or death of members of the public 

Global social media companies are currently under enormous pressure to solve this growing problem as the leaders of the UK, France and Italy have challenged them to remove terrorist propaganda in a timeline as short as under two hours. If they fail to show they are taking action by a G7 meeting on October 20th they will come under the threat of fines, as is the case in Germany, which will see a similar law come into effect this month.

The Crisp Capture project has also identified trends in terms of spikes of new content being uploaded and shared around key events:  

  • Las Vegas Mass Shooting (10/2/17) – Double the number of items uploaded the day following the mass shooting compared to the average amount – the terrorist content was celebrating the attack as a “victory” while encouraging others to do the same
  • Anniversary of 9/11 (9/11/17) – nearly double the daily average of terrorist items were recorded

“Our approach works on the premise that illegal content is being uploaded because the author wants people to see it,” added Hildreth. “Instead of searching through huge quantities of good content in the hope of finding the bad, we monitor at scale the vast data sources that promote bad and illegal content, giving us incredible accuracy, real-time detection and scalability.”

About Crisp

Crisp is the global authority on social media risk and has been protecting hundreds of globally recognized brands and social platforms from the risks posed by social media for over ten years. The world’s largest media and entertainment business; five of the top ten luxury and fashion brands; the largest pharmaceutical company, and the largest broadcaster in the world are just some of Crisp’s clients. 

Our expert Risk Analysts moderate thousands of global social media channels with a combined total of over 1.3 billion fans and deal with three billion pieces of content every month, identifying risks across images, videos and text in over 50 languages.

Crisp is headquarted in the UK with offices in New York, San Francisco, Los Angeles and London.