When joint US-Israel strikes killed Iranian supreme leader Ayatollah Ali Khamenei, the initial shock soon gave way to a deeper unease. At the heart of this anxiety lay the method used in the decapitation of a regional power’s leadership: an accelerated “kill chain” orchestrated with the use of artificial intelligence (Al).

Three years after OpenAl’s ChatGPT brought Al into everyday life, the world now seems to have psychologically accepted the arrival of the algorithmic assassin. Yet, with that acceptance comes a growing recognition that machines are quietly assuming roles once reserved for human judgment in warfare: selecting targets, analysing threats and perhaps one day taking aim.