14 Dec AI at the service of security: the fight against terrorism
1,440 attacks, 7,616 victims: these are the figures telling the story of terrorist attacks around the world from January 1st,2018 to date.
Terrorism is not only defeasible with weapons, but also through technology. Nonetheless, despite ongoing investigations, it is not always possible to combat terrorist organizations in time.
In this regard, Artificial Intelligence is a new and effective tool in the fight against organized crime and terror networks. The automized analysis of big data, the ability to predict the behavior of groups of people and the mapping of the relations of a criminal network, are just a few examples of a new form of fighting illegality that will ideally go as far as predicting possible terrorist attacks.
Predictive analysis correlates data from different sources (social media, surveillance cameras, credit card transactions, online payments, GPS, etc.) and turns them into useful information for security protection.
Although this may seem a simple task at first glance, the diversity of the data to be processed (photographs, video footage, GPS coordinates, invoices, Facebook status, etc.) and the amount of information available make it very complex.
For this reason, predictive analysis is today largely entrusted to Artificial Intelligence systems, which, thanks to their enormous processing capacity, can promptly support the monitoring of huge amounts of data.
A multitude of data
AI systems have the ability to process huge amounts of data that are characterized by heterogeneity and volatility.
Many organizations have made social media their primary channel for political and religious propaganda. According to a recently published report, affiliates of a major criminal organization use more than 50,000 social accounts on various platforms in dozens of different languages to spread the message and recruit new followers. Keeping them under control ‘manually’ is, therefore, impossible.
AI has the task of helping analysts and intelligence services to identify people or groups that manifest potentially dangerous attitudes online: this is done by a real-time analysis of what happens, for instance, on Facebook, Twitter, YouTube, and so on, highlighting suspicious content.
Not only text messages and keywords, but also images, audio files, geographical locations, interactions between people and contents. This is how it is achievable to find out whether a potentially threatening group is building up around a certain account, increasing the chances of giving rise to criminal cells in the future.
AI against narcos
A good example of the application of these systems in support of national security comes from Mexico, where law enforcement agencies have long been using social media analysis to prevent, or attempt to prevent, the bloody attacks organized by drug trafficking cartels in different areas of the country: violent assaults, often resulting in innocent bloodsheds.
As a consequence, it has become common practice in some cities in Mexico to use social media to find out about the security of specific areas while travelling: for example, are there any cars blocking roads (a system often used by hitmen to prevent their targets from escaping)? Or are there any suspicious movements that might presage the worst?
All these pieces of information, tweets, Facebook statuses, and photographs, suitably processed and aggregated by an AI system, can be used by the police as an alert system to keep the territory under control.