While the Terrorism Prevention and Intelligence Bill is under consideration in the Senate, Boaz Ganor, founder and executive director of the International Institute for Counterterrorism (ICT) in Herzliya, north of Tel Aviv, details Israel’s use of algorithms and artificial intelligence (AI) in the fight against terrorism. The main novelty of the law under discussion in France is in fact the perpetuation of the use of algorithms, nicknamed “black boxes”, and their extension to URLs, that is to say to the domain names consulted on Internet.
Since when has Israel been using AI and algorithms in counterterrorism?
In 2015, Israel faced a massive wave of knife attacks, carried out by Palestinian “lone wolves”, which lasted two and a half years. The “lone wolves” are not a new phenomenon in Israel; it started in 1987 during the first Intifada. But in 2015, the security and intelligence services were taken aback by the scale of the phenomenon. This is where they resorted to AI, which proved to be quite effective.
What is the interest of AI against this type of attack?
Among the different types of terrorists, “lone wolves” are individual assailants, who have no operational link with a group. This does not mean that they are not inspired by this or that organization, but they were not recruited and did not receive help to carry out their project.
However, the “lone wolves” do not discuss their plans on the phone or with their families, they do not receive orders. They take action, that’s all. And, among them, there are “spontaneous assailants” who do not even know that they are going to carry out an attack the morning of their act. It is the fruit of an opportunity, a state of mind or an annoyance. The radicalization was already there, but not the project. They are the most difficult to apprehend, because they escape all conventional surveillance, and even their families often do not suspect their radicalization.
It was found that most Palestinian “lone wolves” gave early signals of this radicalization in their interactions on social networks. Unlike common criminals, terrorists act out of “altruism”: they defend their values, their religion or their homeland and let it be known. AI is capable of shuffling a phenomenal amount of data and isolating extremely weak signals, unusual activity or state, that human intelligence would not have spotted. In addition, a study we carried out on “lone wolves” sentenced to prison terms in Israel shows a high level of mental instability. It is possible to seek to detect this kind of behavior online.
You have 64.32% of this article left to read. The rest is for subscribers only.