Gaza City, Gaza Strip – The use of artificial intelligence (AI) in warfare has long been a topic of concern among experts, with fears extending beyond the Hollywood-esque scenarios of autonomous killer robots. Recent reports have shed light on Israel’s utilization of an AI system known as Lavender during its conflict with Hamas in Gaza.
According to Israeli sister publications +972 and Local Call, Lavender has been employed by the Israeli Defense Forces to identify targets for assassination. Unlike traditional practices of assessing targets thoroughly before acting, the system allegedly categorized individuals in Gaza based on various data sources to assign assassination scores. This led to a vast list of potential targets, including as many as 37,000 people at one point.
Despite knowing that the AI system was only 90% accurate in identifying militants, Israeli intelligence reportedly did not conduct thorough human reviews of the targets. This resulted in a high number of individuals being deemed potential assassination targets with little verification, ultimately leading to a significant civilian death toll.
In response to the publication’s findings, the IDF issued a statement denying the use of AI to identify terrorist operatives and ensuring that strikes are carried out in accordance with rules of proportionality. However, the reported use of Lavender may have contributed to the high civilian death toll in the early stages of the conflict, with almost 15,000 Palestinians, primarily women and children, losing their lives.
The revelations surrounding Lavender’s deployment highlight the importance of responsible military use of AI, a topic addressed in a Political Declaration signed by over 50 countries but not including Israel, Russia, or China. As the international community grapples with the implications of AI in warfare, the need for ethical, responsible, and humanitarian-compliant use of such technologies becomes increasingly evident.