Elbit Systems’ drones, using AI and mass surveillance, are at the forefront of “autonomous warfare”, with little or no human oversight

As the Israeli-owned British government and corrupted judges continue their treasonous campaign to imprison, as terrorists, activists taking direct action against factories owned by Elbit Systems, Israel’s biggest arms company, a timely article in Haaretz, published just six days ago, reveals how the real terrorists are Elbit Systems themselves.

In an article on June 5, “’Server in the Sky': How Israeli Drones Became a 'Target and Intel Generator' Over Gaza”, reporter Omer Benjakob revealed how two drones,  the Hermes 450 and the Hermes 900, both manufactured by Elbit Systems, “were deployed at an unprecedented scale in the first months after October 7, 2023”, tracking “nearly every corner of Gaza from the air, identifying and classifying targets for attack.”

The Haaretz article drew on internal IDF documents obtained by the newspaper, revealing “previously unreported artificial intelligence and mass surveillance capabilities embedded into the unmanned aircraft”, which have been “used to help identify targets and direct fire for both the air and ground forces”, and which “lean heavily on AI and underscore the increased role of automation in war today.”

Ground-breaking reports by +972 Magazine, sometimes working with the Guardian, have established how AI-driven military targeting has transformed Israel’s genocidal capabilities, generating targets at a rate that is inconceivable for humans working alone — thousands, or even tens of thousands of time faster. These have then been adopted with very little human oversight, and with no regard whatsoever for an error rate estimated at 10%, or for over-broad definitions of targets in the original programming, which not only include “low-level” targets, often with no military role whatsoever, but also targeting based on alleged behavioral similarities to known militants.

All of this has been used, fundamentally, as an excuse, provided by AI, to indulge in mass slaughter while claiming that it was actually precise targeting — and especially as all notions of limiting civilian “collateral damage” were almost entirely done away with — but there was at least still some trace of human oversight.

What the Haaretz articles makes clear, however, is that Elbit’s drones, which are armed with missiles as well as complex arrays of monitoring equipment, are equipped to make autonomous decisions that make human oversight even more peripheral.

One “previously unreported system”, known as the ‘Server in the Sky’, “running on a computer installed on a drone”, whose “’onboard’ analytics software uses algorithms to carry out a wide range of unmanned missions that utilize AI analysis and decision-making”, “independently analyzes the intelligence gathered by the drones' sensors and cameras, automatically detecting targets, classifying them and deciding whether to track them or pass them on — to the command center, air force pilots or troops on the ground.”

According to Haaretz, the documents reveal that the ‘Server in the Sky’ delivers "AI-based recommendations and decisions” (my emphasis added), which suggests that Elbit is designing and manufacturing weapons in which human oversight may well, in some cases, have been erased entirely.

Haaretz claimed that a human was “clearly in the loop”, but is that true? Assessing the drones’ surveillance system, known as ‘Wide Area Persistent Surveillance’, the newspaper noted that it “has automatic analysis built in, including classification, sorting and ‘gatekeepers’”, adding, “Exactly what that means isn't clear; it might be part of the Server in the Sky's automatic decisions”, and also explaining, “In the world of algorithms, ‘gatekeepers’ refers in part to automatic filters that decide which information gets passed on, and whether to trigger an action or notify a human operator.” (again, my emphasis added).

In response to a query from Haaretz, the IDF stated that, “According to IDF directives, every strike requires an individual assessment by a person regarding the nature of the target, the military advantage relative to the expected collateral damage, and an examination of feasible precautionary measures under the circumstances.”

But, again, is that true, or is the IDF just pretending that human oversight is always present in systems whose capacity is such that they are already working autonomously?

The IDF also stated that they were “committed to the responsible and lawful use of information tools, data and advanced technologies, in accordance with international law”, but that too may well be nothing more than meaningless platitudes offered up to create an illusion of legality.

The truth is that everything about AI-driven military targeting, and the development of autonomous systems capable of operating without human oversight is a legal black hole, an area in which the development of the technology has far outstripped the ability of international legal experts to assess whether or not it is fundamentally legal, or if, as many of us fear, it is a new frontier world of grotesque illegality, designed, as I suggested above, to launder mass extermination as precise targeting.

Worryingly, as the Haaretz article also explains, the perils of this new technology, developed by Elbit, are not just confined to Israel.

According to Elbit themselves, their drones “have been exported to at least 20 countries.” Haaretz added that, “although the company usually keeps its customers' identities under wraps, various publications have reported that some of the systems have been sold to Azerbaijan, Brazil, Mexico, Switzerland, the Philippines, Thailand and Singapore.”

Gaza, it is clear, has been the testing ground for AI-driven military targeting, but now nowhere is safe.

Can Palestine Action really be regarded as terrorists when they seek to disable Elbit’s weapons? Are they not, in fact, on the front line protecting the world from autonomous flying AI-driven weapons systems that inhabit a murderous new frontier of warfare without any accountability or oversight?