Combating online radicalisation with expanded AI capabilities

Building on the work we did on the initial Daesh video classification tool, Faculty has been hard at work expanding both the functionality and adoption of its extremist classification technology. Extended capabilities include the ability to detect both al-Qaeda video propaganda and terrorist still imagery. In parallel, we are working to deploy the tool in the real world to combat terrorism through new partnerships with front-line police forces.

Launched in 2018, the Faculty terrorist classification tool (TCT) is an artificial intelligence (AI) algorithm that is capable of detecting 94% of existing and previously unseen Daesh video propaganda content with a false positive rate of 0.005%. The original purpose of the tool was to work across different video and streaming platforms to flag terrorist content before it is successfully uploaded on to the internet, in order to prevent it from radicalising individuals. This has now been extended to include the ability to detect and flag al-Qaeda video propaganda at comparable accuracy levels.

We are now expanding the adoption of the tool by providing a more efficient and accurate way for the UK’s counter-terrorist police investigators to detect potentially harmful content that might be collected over the course of an investigation. Access to this cutting-edge technology by police forces is helping them to better adapt to new and emerging threats that are created by the rapid rate of technological advancement in this space.

As outlined above, significant progress has been made to identify branded propaganda videos. However, the UK Home Office estimates that more than 80% of content shared online by supporters of terrorism consists of photographs or other still imagery. These images pose a significant threat, as the content is often difficult to detect, analyse and remove owing to its inconsistent or often incomplete branding elements. We are now focusing on creating the technology to detect still image propaganda with similar accuracy.

AI is one of the most important technologies of our time, and can be applied to tackle some of the world’s most pressing challenges. As demonstrated by the work we have done in security, AI is providing significant value in the front-line fight against radicalisation and home-grown terrorism. Building on our work in the security space, we are committed to further developing the tools and technology required to combat the evolving threat of terrorist propaganda.

We are exhibiting at the Security & Counter Terror Expo (SCTX) on 5–6 March 2019 at Olympia, London. Come and visit us at stand C11 to discuss our work or sign up to attend our CEO Marc Warner’s talk ‘Detecting and Countering Extremism in the Online Space’ at 11.45am on 6 March here.