By Bill Taylor/Managing Editor – THE DARK WEB. The very name conjurers up sinister thoughts, mystery, evil and illicit activities. I personally thought it was new Marvel comic book movie. But the dark web is becoming a bit more well lit. First, it’s pretty hard to be an anonymous, secret, deep website if everybody has heard of it. Second, if everybody has heard about it, you can bet tonight’s dinner that law enforcement professionals have too.
So, with police prowling the dark web to catch evil doers and crack down on illegal activity, a new idea thought emerged from an Australian Federal Police (AFP) officer who just happened to be working on a university study. Motive.
To be clear, not all activity on the dark web is illegal. There are things that people just want to keep out of the public eye. Financial transactions (like dark pools) can be privately completed in order to remain anonymous (yes, totally legal) so as not to disclose holdings.
I would name some others but they are, um, secret, ya know? So, how do you (law enforcement) separate legal activity from illegal activity? Artificial Intelligence (AI) is coupled with Machine Learning (ML) to develop algorithms to target illegal material. How? Seems the best way is to analyze the motivation for the activity. Every transaction or post on the dark web has to have a motive and AI and ML can seek that out.
Obviously if Goldman Sachs handles a block trade between two clients who wish to remain anonymous its not illegal. But, if I post credit card numbers I hacked up for sale……..bingo, illegal. Busted. So, the identification of the MOTIVE can be flagged by the AI/ML algos. Studies have shown that 40% of the pages analyzed off the dark web were legitimate. Of course just eliminating almost half of the pages frees up the police to concentrate their crime busting efforts on the bad guys/girls.
So, the dark web is getting lighter (or moved to some other secret venue) and the big finding is? It was not as sinister as thought. Still think it makes a good Marvel movie.
Reference Article: CreateDigital.org.au