On Sunday, Elizabeth Dwoskin reported in the Washington Post, citing two people familiar with the operation, that the use of AI to rapidly refill IDF’s target bank allowed Israel to continue its campaign uninterrupted.
In a June 30, 2023, interview with Ynet, retiring IDF chief of staff Aviv Kochavi revealed that “Among all the technological revolutions, artificial intelligence (AI) is likely to be the most radical, for better or worse. The IDF recognized this field years ago and harnessed it to enhance combat effectiveness.”
Kochavi continued: “One example of this is the Targeting Directorate established three years ago. It is a unit comprising hundreds of officers and soldiers, powered by AI capabilities. It is a machine that processes vast amounts of data faster and more effectively than any human, translating them into actionable targets.
“In Operation Guardian of the Walls, once this machine was activated, it generated 100 new targets every day. To put it in perspective, in the past, we would produce 50 targets in Gaza in a year. Now, this machine created 100 targets in a single day, with 50% of them being attacked.”
Following the devastating Hamas attack on October 7, 2023, the IDF launched extensive airstrikes on Gaza, utilizing a meticulously compiled database containing information on home addresses, tunnels, and other infrastructure deemed critical to the terrorist group.
However, as the initial target database began to dwindle, the IDF relied on the advanced artificial intelligence system that had been described by Kochavi which enabled the rapid generation of hundreds of additional targets, helping sustain the war’s relentless pace.
It’s known as “Ha’Bsora, or “the Gospel.” Not a very Jewish name…
This highlights how a decade-long initiative to integrate advanced AI tools into the core of the Israel Defense Forces’ intelligence operations has played a key role in sustaining the effectiveness of Israel’s 14-month war in Gaza.
Using the software’s image recognition capabilities, soldiers could detect subtle patterns, such as slight changes in years of satellite imagery over Gaza, indicating that Hamas might have buried a rocket launcher or excavated a new tunnel on agricultural land. This technology compressed what would have taken a week of analysis into just 30 minutes.
Another machine learning tool, known as Lavender, assigns a percentage score to predict the likelihood that a Gaza Arab is a member of a terrorist group, enabling the IDF to rapidly generate a large number of potential human targets. Other algorithmic programs include Alchemist, Depth of Wisdom, Hunter, and Flow—the last of which, previously unreported, allows soldiers to query multiple datasets efficiently.
During the early stages of the Gaza war, the IDF granted broad authorization for officers to adopt Lavender’s kill lists without requiring a thorough review of why the AI system selected those targets or an examination of the raw intelligence data behind the decisions. According to one source, human oversight often amounted to little more than a “rubber stamp,” with personnel typically spending just “20 seconds” per target—mainly to confirm that the Lavender-marked individual was male—before approving a bombing. However, even the IDF detractors, such as the anti-Israeli website 972, acknowledged that the error rate of Lavender was only around 10%.
It should be noted that the IDF has learned a thing or two from its enemies and systematically targeted wanted terrorists in their homes, at night, when their entire families were present. This means the rules of engagement have been changed radically, resulting in much fewer casualties on our side.
According to 972, one AI system called “Where’s Daddy?” was specifically designed to track targets and facilitate strikes when they returned to their homes.
The result of these AI-driven targeting systems was the deaths of thousands of Arabs, at least 20,000 of them were confirmed terrorists, while the IDF losses in the post-October 7 months have only been numbered in the hundreds. This devastating impact on the Gaza Arab side was particularly pronounced during the initial weeks of the war, as Israeli airstrikes heavily relied on decisions generated by the AI programs.
