Moral responsibility for AI systems must rest with the humans who build and deploy them, not with the machines; the risk is that humans will abdicate moral accountability by claiming the machine made the choice, particularly in military contexts where autonomous weapons could commit atrocities without human culpability.