AI warfare is already here
anthropic's dispute with the pentagon highlights the risks of autonomous warfare, obscuring how close it is.
anthropic currently resists pentagon demands by enforcing two bans: one on domestic mass surveillance and another on weapons that identify and kill targets without human intervention. this stance contrasts sharply with the history of project maven, a 2017 initiative that used google ai to analyze drone footage and automate target selection. the un convention on certain conventional weapons shifted from hypotheticals to reality that year, acknowledging that autonomous platforms were already in development. experts warn that despite anthropic's limits, the military has embedded ai deeply into warfare for seventy years to accelerate killing efficiency.