Israeli authorities have foiled over 200 Palestinian attacks by monitoring social media and sifting through vast amounts of data to identify prospective terrorists ahead of time, said Public Security Minister Gilad Erdan.
These pre-emptive actions put Israel at the forefront of an increasingly popular — and controversial — trend used by intelligence and law enforcement agencies around the world that use big data technology to track would-be criminals. While the technology appears to be effective, its tactics drew angry Palestinian condemnation and have raised questions about civil liberties.
Erdan, who oversees the national police force, said Israel’s use of algorithms and other technology has been an important factor in lowering the number of knife and shooting attacks in Israel in recent years. He planned on sharing Israel’s knowledge with counterparts at an international security conference he is hosting that begins Tuesday.
“The experience we now have, we can help other countries deal with this kind of terrorism,” he said. He said working with allies “can lead us to a much better result in fighting lone wolf terrorists.”
In September 2015, Israel found itself facing a wave of stabbings, shootings and car rammings carried out by “lone wolf” attackers, or individuals unaffiliated with militant groups acting on their own. It was a significant departure from past waves of organized violence led by armed groups like Hamas.
Since then, Palestinians have killed over 50 Israelis, Hy”d, while Israeli forces have killed over 260 Palestinians, most of whom terrorists. However, the number of attacks has dropped significantly — from 170 “serious attacks” in 2016 to 90 last year to 25 this year, according to Erdan’s ministry.
Israel has blamed the attacks on anti-Israel incitement in Palestinian social media.
Erdan, who is also minister of strategic affairs, the agency responsible for fighting the BDS movement, said Israel has turned to various technologies to counter the attacks. That includes facial recognition devices and smart cameras that detect suspicious behavior in real time.
In addition, his ministry, working with the Justice Ministry and Shin Bet internal security agency, has created a team to scour an “ocean of data” on social media for objectionable content and to identify attackers before they act.
Members include psychologists, legal advisers and experts who have developed algorithms that analyze online activity. Violent posts, the suspect’s profile, such as age or hometown, and other supporting evidence are factored into the analysis.
“Every event can lead to a discussion. You have to look for the special words that might lead you to the conclusion that something is dangerous,” Erdan said. “The algorithm leads you to suspect someone.”
The system has nabbed over 200 people who Erdan said confessed that they were planning attacks. Suspects have included both Palestinians and Arab citizens of Israel, including cells allegedly inspired by the Islamic State terror group.
“When you increase the number of people who are involved in monitoring the social networks, you more and more learn how to define the profiles you are looking for,” he said.
Erdan said there is always a “dilemma” about whether a potential suspect is a serious threat or not, and that police often have just a few hours to decide whether to make the arrest.
“It’s complicated,” he said.
There were no figures on how many innocent people were misidentified as potential suspects. But Erdan’s office said in such cases, the people are quickly released.