Jun 1, 2019

Automated war

A Yemeni man looks at graffiti protesting against US drone strikes on September 19, 2018 in Sana'a, Yemen.

A mural in Sana’a, Yemen protesting against U.S. drone strikes. Photo: Mohammed Hamoud/Getty Images

A new era where weapons of war are becoming more intelligent and more enabled by data — such as unmanned ships, submarines or drones — raises complex challenges for national and global security.

Threat level: If technology is allowed "to start making big decisions on its own ... we might be doomed by technological advances," David Petraeus, former CIA director and retired four-star general, tells "Axios on HBO."

Driving the news: Experts are grappling with the ethics of developing autonomous weapons, which suggest the possibility of a computer deciding on its own to take human life.

  • UN Secretary-General António Guterres has urged AI experts to ban autonomous weapons, calling them “morally repugnant.”
  • Drones — low profile and easily preprogrammed with GPS routes — are just the beginning of warfare with AI. Recently, a drone blast killed several people in Yemen, including the Yemeni government's head of military intelligence.

Fully autonomous weapons don't exist today, but Petraeus warns even a world of semi-autonomous weapons could create a frightening future for humankind.

  • "All of this is advancing so rapidly that it’s literally difficult to keep up with it intellectually conceptually."

Our thought bubble, per Axios AI reporter Kaveh Waddell: The big challenge now is to slow the world’s slide toward an automated weapons race fueled by mutual distrust and a lack of information.

  • As international efforts to ban autonomous weapons stall — in part thanks to the U.S. — look for the Pentagon to update its policies on automation late this year.

Go deeper:

Go deeper