Behind NAZA: How AI is used to assess, reduce collateral damage in IDF - interview

The article examines the controversy surrounding the documentary 'NAZA,' which alleges that IDF AI systems facilitate mass killing in Gaza. It contrasts the film's claims with the IDF's position that AI tools like Lavender and Habsora are used to assess and minimize collateral damage, not to automate targeting. The piece highlights the distinction between calculating civilian harm and the quality of that calculation, noting disputes over permitted thresholds and human oversight.

The article examines the controversy surrounding the documentary 'NAZA,' which premiered at the Venice Film Festival and received the Special Jury Prize. The film, by Israeli filmmakers Yuval Abraham and Rachel Szor, alleges that IDF AI systems facilitate what it describes as 'calculated mass killing' of Palestinian civilians in Gaza. The article contrasts this framing with the IDF's position that AI tools like Lavender and Habsora are used to assess and minimize collateral damage, not to automate targeting. Cyber intelligence expert Matías Rott is quoted explaining that AI serves as a decision-support function to reduce civilian harm. The piece highlights the distinction between calculating civilian harm and the quality of that calculation, noting disputes over permitted thresholds and human oversight. It references Abraham's own 2024 reporting for +972 Magazine, which described collateral-damage calculations as a condition for strikes but also alleged that permitted thresholds were too high during the early war. The IDF has rejected the film's allegations and challenged the reliance on anonymous witnesses.