Skip to content
eBook Charts
US
Robots in Command

Description

An urgent warning that growing reliance on autonomous weapons will force human commanders to yield battlefield decision-making to AI, increasing the risk of civilian casualties and nuclear escalation.

The robot army is already here, and it’s only gaining in capability and sophistication. Soon, US soldiers, sailors, and pilots will give way to unmanned weapons systems and vast swarms of self-directed combat drones. China and Russia are pursuing similar strategies, each seeking a technological edge over its rivals.

The Pentagon insists there will always be a human “in the loop,” making decisions involving the use of force. But, as Michael Klare explains, these assurances offer little comfort. On future battlefields, multitudes of robotic sensors on the front lines will unleash torrents of information, overwhelming commanders. The only viable option, Pentagon officials claim, will be to rely on artificial intelligence for data synthesis and battlefield management, leading to ever-diminishing human oversight. And as AI systems react to each other’s moves at superhuman speed, even a slight computer error or sensor malfunction could cause a conflict to spiral out of control, resulting in a nuclear “flash war” analogous to the flash crashes already seen in Wall Street’s high-frequency stock trading.

We can avert some of the danger by inserting “circuit breakers” into autonomous weapons, designed to halt operations when systems exceed predetermined bounds or lose contact with human managers. The technology to build such safeguards is within reach; what’s needed is a formal requirement that it be installed in all AI-governed devices developed for military use. Robots in Command argues that there may be no stopping AI’s battlefield takeover, but there are critical steps we can take right now to avoid catastrophe.