AI military officers could be making battlefield decisions by 2030, says former CIA chief
AI military officers could be making decisions on the battlefield within just four years, the former head of the CIA has said amid growing debate over the use of the technology in war.
Ret. Gen. David Petraeus said he believed that warfare would begin to see the adoption of AI staff officers that are capable of gathering information and making decisions with little to no human input by 2030.
“Just four years from now, we will have autonomous systems, and I think we will have the beginning of autonomous systems of autonomous systems; we will have AI agents, so you’ll have agentic AI staff officers,” he told investor Finn Murphy on the Forecast2050 podcast, to be released on Thursday.
He also cited the potential of recursive learning, when an AI model uses its own data and processes to improve itself.
”All of that, just again in three and a half years is going to be truly revolutionary. We’d be much better off if we were anticipating more of that,” the former head of US Central Command said in comments shared exclusively with The Independent ahead of publication.
There is skepticism over whether AI could replace the traditional military chain of command. A study published last year by the Alan Turing Institute found that AI could assist commanders and improve the speed of decision-making, but not replace human judgment or accountability for life-or-death decisions.
The former four-star general, who served under president Barack Obama from 2011-2012, described such systems as “generic artificial intelligence staff officers who will never sleep, never eat, never take a break”.
He has previously said he believes that autonomous systems will “very soon” be working together in platoon or battalion-sized units, able to share information and coordinate without human intervention, adding: “The side that waits for human approval before acting will lose.”
Speaking to CNBC in May, he described ‘autonomous systems of autonomous systems’, where sensors would be able to collect battlefield data and relay it to an autonomous HQ, which could then direct autonomous weapons with little to no human input.
The former intelligence chief told Forecast2050 added that he expected to see self-governing drones being used in the war in Ukraine by the end of this year.
“We’ll see this emerging in Ukraine I think by the end of this year, very likely by the end of 2026. You’ll see increasing numbers of autonomous unmanned systems. And the reason that’s so dramatic is because it now eliminates the limit that you have on the number of systems based on the number of pilots that you have.”
The early adoption of artificial intelligence in warfare remains controversial. US Senate Democrats wrote to defence secretary Pete Hegseth in March seeking information about a reported strike on a girls’ school in southern Iran and requesting to know whether AI was used in planning or executing an attack.
Admiral Brad Cooper, the head of US Central Command, has confirmed the US military has used AI to sift through large amounts of data, without detailing specifics. He said human operatives will “always make final decisions on what to shoot and what not to shoot, and when to shoot”.
Petraeus also said the UK was well positioned to become a world leader in autonomous warfare in spite of its financial woes, while the US is still missing the elements of institutional change needed to transition.
