Criminal intelligence process focuses on the collection of data and information relevant for police investigation and criminal procedures. In many cases information gathered cannot be immediately and directly used as evidence in criminal procedures due to insufficient level of certainty and reliability of the source. It is legally and technically easy for the police services to exchange information of high certainty, but the problem arises in cases of low certainty of information gathered in early stage of investigation and whether it would lead to successful prevention, detection or investigation of the crime committed and to apprehension of perpetrators.
Police services handle this type of information with special care, as it is usually classified as “soft information” or “raw intelligence”. In many instances there are legal obstacles for a direct exchange of this type of information internationally. Services are hesitant to reveal their intelligence to each other as they cannot be sure on the quality of information. They also fear the possibility that sensitive intelligence data will not be used in a proper way. For this reason, lots of available information is never exchanged or at least compared among partner services. “Aquila” focused on the ways to overcome this problem in a way allowing partner services to learn of the fact that they have some matching information in their respective intelligence repositories, whereby all of their data remains securely hidden (“data matching”). After the “hit” occurs, they can proceed with the exchange based on existing international legal bases.
Criminal police units dealing with counter terrorism and extremism-CVE of the PCC SEE region have decided to initiate the process of data matching based on comparison of their national lists of subjects, possibly linked to terrorism or violent extremism. They did not reveal any unmatched data to each other.
Aquila was implemented in the form of three consecutive workshops during 2018. In Sarajevo in May 2018, our experts presented the basic principles of the data matching process and requirements that had to be met by participating parties. Their main task was to prepare a list of subjects’ names and subjects’ electronic ID’s used in cyberspace communication. At the second workshop in Ljubljana in September 2018, national data lists were matched one against another. Out of total 1.385 subject’s records 38 were matched and out of total 388 subject’s electronic ID’s 1 was matched. Participants agreed to include more of their own data for the third workshop, which took place in Ljubljana in November 2018. On this occasion, 183 out of total 1.925 subjects’ records were matched and 3 out of total 690 subjects’ electronic ID ’s were matched.
DCAF engaged IT expertise to create proprietary software in order to automatize the matching process. We installed the Aquila software package on laptop computers of the participants. We then connected all computers to wire-connected ad hoc local area network and established connections among computers. After initiation, each national dataset has been computed on respective laptops in the way, that all records were unified in terms of literation, according to ICAO standard. In the next step, all possible permutations of names (consisting up to five elements of the name) were calculated, and finally all records were coded applying RSA algorithm, making impossible for humans to directly read and understand the information. After all computers reported ready, the essential part of procedure called Private Set Intersection was initiated. Computers started to compare record by record with records on the lists of other computers. After the conclusion of 90 matching processes, all computers reported of the successful conclusion of the procedure. Aquila created a list of matched data for each participating party separately. Records were only revealed to parties with resulting matches/hits. No central database has been created during the process.
In the weeks following the last workshop, participants were requested to continue the process on a bilateral basis, with what they gained with the help of Aquila, in order to improve their “intelligence picture”. We anticipated that there was a certain possibility that those parties, which have successfully matched basic data on some subjects, would be in possession of different information on the same subjects. All combined, this particular information will improve the knowledge level and provide for additional planning of the common intelligence process between the parties.
Participants agreed that the Aquila methodology and software solutions could be used on any type of data and in any area of transborder and international crime. Participants would also welcome the information technology solution, which would be available 24/7 and would not require participants to meet face-to-face in order to execute the matching process.