International negotiations in Geneva to restrict fully autonomous weapons faced setbacks after delegates from the U.S. and Russia weakened proposed treaty guidelines, according to a report by Kenneth Roth. The changes diluted provisions requiring human oversight for military targets identified by AI systems, limiting international focus primarily to fully autonomous “killer robots” rather than AI-assisted targeting tools.
The report highlights current military uses of AI targeting, citing the Israeli military’s deployment of systems named “Lavender” and “Where’s Daddy?” in Gaza to identify suspected operatives and track them to their homes. While human intelligence officers reviewed targets, the process functioned largely with minimal meaningful human control. Similarly, armed forces in Ukraine are testing autonomous drone navigation to counter signal jamming.
Human rights advocates warn that the erosion of meaningful human control in targeting decisions sets a dangerous precedent for modern warfare and political repression. Existing treaty discussions remain focused on wartime scenarios while failing to address domestic surveillance and autonomous force against civilian populations.
Why it matters
Policy makers and defense contractors face mounting international pressure regarding meaningful human control standards in military AI systems.
Diluted global regulations signal a shifting legal landscape for AI target selection, autonomous drones, and military hardware.
Source: theguardian.com



