According to various army publications and journalistic investigations, the military employs artificial intelligence systems in the context of the war in the Gaza Strip. According to the reports, one system, called "HASBORAH," is capable of generating attack targets at a high rate using an algorithm that extracts intelligence data from a wide range of sources. The system processes the information and proposes attack targets, which are then approved by a human authority based on the system's recommendations. The army disclosed its use of artificial intelligence two years ago, and in the current war, it updated the public that it relies on the "HASBORAH" system.
On December 10, 2023, we sent legal correspondence (Hebrew) to the Chief Military Prosecutor and filed a Freedom of Information request (Hebrew) regarding the use of the system. Attorney Gil Gan-Mor, Director of the Civil and Social Rights Units, highlighted the difficulty in relying on artificial intelligence systems for making sensitive decisions that involve severe violations of human rights. Among other concerns, artificial intelligence systems are prone to biases, and their "conclusions" are received as a black box, without a human authority having the ability to monitor the system's operation and justify its recommendations.
Despite this, we requested to disclose all relevant information about the "HASBORAH" system that can be made public. From the Chief Military Prosecutor, we requested an urgent examination of the system and the use of artificial intelligence for decision-making regarding attack targets. We specifically asked to examine how the system aligns with the obligations imposed on the IDF by both Israeli and international law. We also inquired if there are any biases that may lead to incorrect targeting and cause severe harm to uninvolved civilians.
