Just a few stories about how our Open Source Intelligence investigations have helped others
Today lots of organisations use our Actionable Intelligence Services to improve investigations and make the world a better place.
TERRORISM AND RADICALISATION
A radical video showcasing a radical imam’s speech had been uploaded to various social media platforms. The speech promoted violent action against unbelievers and went radical in a matter of minutes. Our client, a Law Enforcement Agency, wanted to identify where the video came from and how it was distributed on social media.
As the video was mostly shared on Twitter, we primarily focused on this social media platform. Having the URL of the video, we were using this as token to search Twitter for posts including this link. Based on these posts, we were then able to identify those who were retweeting the content – allowing us to build the network of users in contact with the upload. By tracing back the retweets and discussions, we were able to then identify the user who originally posted the video as well as the one who had injected the video to Twitter where the content went viral.
ONLINE CRIME DETECTION
We were asked to participate in this interesting yet challenging project when the client, a Law Enforcement Agency’s antinarcotics department had a covert agent infiltrated into a private WhatsApp group. This group acted as a communication channel between drug dealers. Here they retrieved information regarding the timing, whereabouts and other information of the drug handover. However, this important information was cyphered, using special terms, which were changed frequently to guarantee the “safety” of the actors in the business.
Our task was to decipher and understand the details of the next transaction. For deciphering and understanding the dynamic corpus we used our expertise in Natural Language Processing. Having revealed what the encoded messages were telling to the dealers, the forces at the antinarcotics department consequently raided the premises and detained the participants.
HARASSMENT ON SOCIAL MEDIA
Our client, a VIP based in a country with high Twitter penetration, wanted to understand whether attacks on this social media platform against our client were orchestrated in a top-down manner or instead issued spontaneously. The story started with our client making a self-admittedly, all be it rather unfortunate, comment on someone who had many social media followers and fans. Our client identified the comment as being unnecessary, but did not imagine that it would have such a dramatic impact and consequence as fans began to launch vicious attacks on Twitter.
Using Social Network Analysis on the tweets related to our client, we were able to identify the key actors in the discussions. We started by reviewing what these users were talking about, what topics they mentioned along with our client and what other issues were raised. This was completed by using Natural Language Processing. From this analysis, we identified that the very same topics were used by the key actors, and by then analysing their follower network it turned out that there were overlaps between them. This suggested that they belonged to the same (maybe even artificial) circle – suggesting that the attack was indeed orchestrated.
Our client, a national Law Enforcement Agency, had the suspicion that terrorist activity was financed by domestic banks in their respective country and turned to us for validation. We received an anonymised dataset containing information regarding transaction type, location, store details and the total cost.
We assumed that finding the fraudulent transactions would require special anomalous demeanour, and so we started to look for two things in the dataset primarily; outliers and irregular patterns. Additionally, using the bank account holders’ anonymised IDs, we were also able to include the network analysis tools within the model we built. Using Machine Learning algorithms, we were able to differentiate and group the transactions. The weekend shopping, the family eat-outs, the seasonality expenditures – such as Christmas or Valentine’s Day gifts- were clearly distinguishable. We also detected a cluster of transactions, which were significantly different based on several features – one of them being network topology.