Israel is using an algorithm to spot the next ‘lone wolf terrorist’, and soon other countries will too
Fadi Zebar stands in front of a mural in Israel depicting his brother. (ABC News: Eric Tlozek)
Fadi Zebar fits the digital profile of a potential “lone wolf terrorist”.
- In 2016 Israel’s military promoted the use of “predictive intelligence” technology
- Israel claims to have prevented 590 attacks in the past year
- The US is developing a similar system
He’s a young Palestinian man from a restive village in the West Bank, where houses have been demolished and people jailed for killing Israelis.
His brother Imad had been jailed, allegedly for throwing stones at soldiers.
But Fadi, a 27-year-old electrician who just got married, swears he did not have violence on his mind when he posted comments on Facebook about an Israeli raid on his home and shared a video of a protest which also praised a man who attacked Israelis.
Fadi with a picture of his brother Imad, who was jailed by Israeli authorities. (ABC News: Eric Tlozek)
“I posted a video clip of an event that took place in our village and I got many views and likes and [the Israelis] didn’t like that, they got excited,” he said.
“Because of that I was arrested — not only for this reason — but I was interrogated for writing, ‘Good morning martyrs’ … and many more simple things.
“[Israeli soldiers] came to arrest me, they destroyed the main door of our house … I was taken for interrogation, then I was sentenced to serve time in military prison.”
Fadi’s age, the location of his home, his brother’s arrest and his social media posts are likely to have flagged him to Israeli military intelligence, who monitor many Palestinians’ communications and use a computer algorithm to help determine whether they could be planning violence.
In 2016, Israel’s military promoted the use of so-called “predictive intelligence” technology to combat a wave of spontaneous or “lone wolf” terrorist attacks by Palestinians that began the previous year.
Now, the rate of violence has dramatically decreased and Israel’s internal security service claimed to have prevented 590 attacks in the past year.
Fadi was not charged with planning an attack, but he spent nine months in jail for the offence of incitement and assisting a terrorist organisation.
“I was told I was creating a big threat. After my arrest that danger was no longer there, the level of violence dropped down,” he said.
“It sounded like I was the one who caused all of the problems in my village.”
Data collection and a ‘scoring system’
The director of the Military and Strategic Affairs Program and Cyber Security Program at Israel’s Institute for National Security Studies, Gabi Siboni, said software helps Israel’s military decide who might pose a threat.
“We have the technology to try to collect information and to provide some kind of a scoring system to see whether someone is subject to conducting what we call a spontaneous or influenced terror attack,” he said.
“It’s not the only thing.
“There is human intelligence, there is a variety of technological intelligence, gathering other intelligence … the whole effort is to foil these spontaneous attacks (and) it’s a huge effort.”
But human rights lawyer Sahar Francis said security agencies are not just focusing on potential attackers, but also on anyone disseminating political material, for fear it might inspire other attacks.
Sahar Francis says the number of cases based on social media posts is increasing. (ABC News: Eric Tlozek)
“Any activity … posting things on Facebook or in Twitter or other tools in social media — sharing posts of martyrs, posts of activists — like especially this year concerning the demonstrations around Gaza,” she said.
“These activities would be considered incitement and people would be arrested for such activities, facing sentences of a couple of months in prison.
“There’s more than maybe 300 or 400 cases in the last three years were submitted based in such activities. And we feel the increase is still going on.”
Because the West Bank is under Israeli occupation, its Palestinian residents are tried in military courts, where Ms Francis said they are often not given access to intelligence material used to prosecute them.
The risk of ‘false positives’
Former military intelligence officers have already raised concerns about the level of electronic surveillance of Palestinians, saying there were no legal limitations protecting their privacy.
One former officer, who is not allowed to be identified, told the ABC the use of software to process surveillance data and help determine potential attackers would inevitably lead to injustice.
Former military intelligence officer “Lieutenant G” is concerned about the lack of privacy and legal protections for Palestinians under surveillance. (ABC News: Eric Tlozek)
“It’s very clear that arresting innocent people is bound to happen by this method and it already happens,” he said.
“It’s not like this is a new thing, it just gets more extreme.”
The developer of some of the Israeli software, Fifth Dimension, has acknowledged the system will produce “false positives”, meaning it will mistakenly flag some people who pose no threat.
But Gabi Siboni, from the Institute of National Security Studies, said the technology will not lead to people being arrested for crimes they have yet to commit.
“It doesn’t mean that you’re going on and arresting people who you think are going to do something,” he said.
“This is very futuristic and not realistic.”
Israel’s internal security agency and the Israel Defence Force did not respond to questions about the technology, how extensive and effective it was and whether there were any safeguards around its use.
The use of predictive intelligence has been trialled by some police departments in the US and Europe.
The United States is developing a system that works like Israel’s, to determine whether someone is a security risk when they fly, a system that will be offered to other countries to use as well.
So in more places around the world, computers will be deciding who might be dangerous.