‘The machine did it coldly’: Israel used AI to identify 37,000 Hamas targets

Key Points:

  • Israeli military used AI-powered database to identify potential targets in Gaza
  • Israeli officials permitted large numbers of Palestinian civilians to be killed during the conflict
  • Israel’s use of AI in warfare raises legal and moral questions about targeting and collateral damage


The Israeli military’s recent bombing campaign in Gaza utilized an undisclosed AI-powered database called Lavender, which identified 37,000 potential targets tied to Hamas, claim intelligence sources. This revelation sheds light on the military’s use of machine-learning systems to detect targets during the six-month conflict, pushing the boundaries of advanced warfare and sparking legal and ethical concerns about human-machine interaction on the battlefield.


The candid testimonies of six intelligence officers paint a picture of how Lavender played a pivotal role in rapidly pinpointing Hamas and Palestinian Islamic Jihad targets for airstrikes. The officers revealed that the system, developed by the Israel Defense Forces’ Unit 8200, led to a staggering number of potential target identifications, particularly lower-ranking Hamas affiliates, amidst the escalating violence.


Noteworthy details involve the use of pre-approved casualty thresholds for civilian deaths in airstrikes, with officers indicating permissions to target individuals regardless of rank, even if it meant significant collateral damage. This approach highlights a shift towards a more radically permissive stance on civilian casualties during the conflict. The intelligence officers voiced concerns over the increasingly aggressive targeting and the consequences of such strategies, acknowledging the toll on civilian lives amid the escalation of the conflict.
The methodology of Lavender’s data analysis, its impact on target selection, and the debate over proportionality in assessing collateral damage are key focal points. Sources describe how the system’s algorithm evolved and the challenges of defining targets amidst fluctuating casualty thresholds, shedding light on the ethical dilemmas faced by military personnel using AI-driven systems.


The testimonies raise broader international law concerns regarding proportionality and civilian protection in armed conflicts, with experts voicing alarm over the reported high ratios of permissible civilian casualties, particularly concerning lower-ranking combatants. The testimonies spark questions about the moral and legal justifications behind Israel’s aerial bombing tactics and the long-term implications for civilian populations in the conflict zone.


The revealed insights, shared prior to an investigative report’s publication, offer a critical examination of the impact of AI in modern warfare settings and challenge the traditional notions of target selection, civilian protection, and ethical decision-making in armed conflicts. The testimonies underscore the complex interplay between AI technologies, military strategies, and the human cost of conflicts, calling into question the balance between military objectives and civilian harm in contemporary warfare scenarios.



Prompt Engineering Guides



©2024 The Horizon