War in the Gaza Strip: Israel uses AI systems to identify targets – News


Contents

Research shows how heavily Israel relies on artificial intelligence in its attacks: Will Gaza become a laboratory for the war of the future?

There has been war in the Gaza Strip for six months. Israel reacted harshly to the Hamas massacre on October 7th – and has never made a secret of it. Even in the first days of the war, the head of the Israeli Air Force spoke of “relentless air strikes – around the clock.” And added: “We are not proceeding surgically.”

The killing of foreign aid workers and the high number of civilian casualties in Gaza are causing increasing criticism of Israel’s conduct of the war. Even the USA, Israel’s most important ally, is raising the warning finger. Israel emphasizes that aid organization employees were targeted because of false identification. But how does Israel choose its targets in the Gaza war?

Netanyahu: Number of civilian casualties decreased


Open the box
Close the box

Legend:

According to Israeli Prime Minister Benjamin Netanyahu (pictured), civilian casualties are inevitable.

Keystone/EPA/AP/OHAD ZWIGENBERG

Israeli Prime Minister Benjamin Netanyahu recently stated that civilian casualties were inevitable in the war in Gaza. But the army is trying to keep the number as low as possible. In the Interview with the German war reporter Paul Ronzheimer Netanyahu said that Israel’s army had killed “at least 13,000 terrorists” in Gaza. “The number of civilian casualties was probably a ratio of 1.5 to 1. Now it is only 1 to 1.”

The Israeli head of government was referring to the ratio between terrorists killed and civilians killed – and that more civilians died at the beginning due to intensive air strikes than in the current phase with more ground operations. At the same time, Netanyahu accused the Western public and politicians of double standards: In wars like Syria, Yemen and Iraq, the civilian death toll was much higher than in Gaza – but there was no outcry about it.

According to the Hamas-controlled Palestinian authorities, 32,975 people have been killed in the Gaza Strip since the Israeli offensive began. Around 75,577 people were injured (as of April 3rd).

According to Research by the British “Guardian” Israel is increasingly relying on systems that use artificial intelligence. The focus is on the AI ​​platform “The Gospel”. It is used by the Israeli military to quickly analyze large amounts of data and identify potential attack targets.

Israeli airstrike in the Gaza Strip on May 13, 2021.

Legend:

The Israeli army prides itself on its technological expertise. During their eleven-day intervention in the Gaza Strip in May 2021, high-ranking military officials spoke of the “first AI war”.

Keystone/AP/Hatem Moussa

“These targets can be individual fighters, equipment such as rocket launchers or facilities such as Hamas command posts,” explains Anselm Küsters. He heads the Department of Digitalization and New Technologies at the Center for European Politics in Berlin. Kusters has delved deeper into the use of AI in wars.

Huge gain in efficiency through the use of AI


Open the box
Close the box

For example, AI systems can aggregate information from social media, help with target targeting, or guide troops and vehicles through the war zone. According to Küsters, Israel also uses facial recognition programs to identify suspects. The AI ​​programs often came from start-ups that specialized in specific areas of information gathering.

Huge amounts of data are collected. According to the expert, Israeli AI systems are probably based not only on cell phone surveillance, but also on satellite images, drone footage and even seismic sensors. They combine to form an overall picture of where individuals and groups are presumably located.

According to the British “Guardian”, Israeli military officials are said to compare “Gospel” to a factory that continually produces potential targets for attack. Kusters illustrates this deadly efficiency: Previously, the Israeli military relied on large groups of officials who suggested around 50 to 100 targets within a year. The AI ​​platform “The Gospel”, on the other hand, can identify 200 targets within 10 to 12 days.

An AI that independently pursues human goals and recommends them for elimination: That sounds as futuristic as it is terrifying. “This war will be a milestone if the Israeli army uses AI in a significant way to decide between life and death,” a former White House security adviser told the Guardian. “Other states are looking at Gaza and will learn.”

The high number of victims in the Gaza Strip is also due to the fact that Hamas abuses civilians as human shields, Kusters makes clear: “The use of AI by Israel must also be understood against the background of the massacre on October 7th. The threat from Hamas should be eliminated and hostages should be rescued.”

It is a misconception that a combination of people and machines is sufficient for a safe application.

At the same time, he is critical of the use of AI for killing purposes. AI algorithms are “notoriously error-prone”: They deliver results based on statistics and historical data – and these are often not accurate in rapidly changing contexts such as war. “There are therefore good reasons to doubt whether an AI will ever be able to differentiate between military and civilian with 100% certainty.”

Will the AI ​​soon kill itself?

The AI ​​systems that Israel uses do not operate autonomously: In the end, a human decides whether a target is attacked. However, the sheer amount of data that AI makes available can also put military personnel under pressure – and lead to blind trust in the technology, says Küsters: “It is a misconception that a combination of humans and machines is a safe application is enough.”

Pro-Russian fighter controls demining robot in Ukraine.

Legend:

For Küsters, the war in Gaza and Ukraine seem like “tragic laboratories” on the way to AI that kills itself. Image: Pro-Russian fighter controls demining robots in Ukraine.

Reuters

Experts warn of a horror scenario: AI weapon systems that also kill themselves. For Küsters, this could soon become a sad reality: “We are approaching a new phase of war in which Gaza and Ukraine act as a test laboratory.”

Struggle to ban “killer robots”


Open the box
Close the box

“Killer robots” are no longer science fiction: countries like the USA, China and Russia are already experimenting with autonomous systems such as swarms of drones. The regulation of autonomous AI weapon systems will be a major challenge for the international community. Civil society organizations and the UN have been trying for years to introduce binding bans worldwide.

In the current “geopolitical race for technology,” expert Küsters currently sees little chance of success for a global consensus. But this is urgently needed: “One of the central characteristics of these AI systems is that they learn themselves and can generate themselves to a certain extent autonomously. The most effective way to rule this out from the start would be some form of ostracism.”

source site-72