Autonomous weapons proliferation reflects militarised AI governance failures, not inevitability
Original framing: “Killer robots are here – we must finally decide whether to accept them” — New Scientist
Scientific consensus warns that autonomous weapons lack the contextual understanding to comply with international humanitarian law, particularly in distinguishing combatants from civilians in complex environments. Studies from MIT and Stanford highlight the unreliability of AI in high-stakes decision-making, with error rates exceeding human judgment in stress scenarios. The 'kill chain' of autonomous systems also introduces cascading failures, where a single misclassification can trigger disproportionate violence, as seen in drone warfare.
The autonomous weapons crisis is not a technological inevitability but a failure of governance, where decades of unchecked militarisation of AI have been legitimised by a discourse that frames ethical dilemmas as 'choices' between dystopia and acceptance.
This framing serves the interests of defence contractors, Silicon Valley elites, and state militaries, while obscuring the disproportionate harm to marginalised communities in the Global South, where Western arms exports fuel proxy wars. Historical patterns—from the machine gun to nuclear arms—show that unregulated weaponisation leads to escalation and catastrophe, yet the current debate repeats these mistakes by treating autonomy as a neutral 'capability' rather than a moral abomination. Indigenous epistemologies and Global South philosophies offer alternative frameworks that reject the commodification of violence, but these voices are sidelined in favour of a technocratic monoculture. The path forward requires binding international moratoria, democratic control of military AI, and a radical reallocation of resources toward peace technologies—all grounded in reparative justice and the wisdom of those most affected by war.