The Pentagon used Project Maven-developed AI to identify air strike targets

The US military has increased its use of artificial intelligence tools since the October 7 Hamas attack on Israel. Bloomberg. Schuyler Moore, chief technology officer at US Central Command, told the news organization that machine learning algorithms helped the Pentagon identify targets for more than 85 airstrikes in the Middle East this month.

U.S. bombers and fighter jets carried out airstrikes on seven targets in Iraq and Syria on February 2, destroying or at least damaging missiles, rockets, drone depots and militia operations centers. The Pentagon also used artificial intelligence systems to locate missile launchers in Yemen and surface combatants in the Red Sea, then destroyed them in multiple airstrikes that month.

The machine learning algorithms used to narrow down the targets were developed under Project Maven, Google’s now-defunct Pentagon partnership. Specifically, the project involved the US military using Google’s artificial intelligence technology to analyze drone footage and flag images for later human review. This caused an uproar among Googlers: There were thousands of them petitioned the company ended its partnership with the Pentagon, and some even refused to participate altogether. A few months after that employee protest, Google made a decision not refresh His contract expires in 2019.

Moore said Bloomberg US forces in the Middle East have not stopped experimenting with using algorithms to identify potential targets using drone or satellite imagery, even after Google’s involvement. The military has been testing their use in digital training for the past year, he said, but began using the targeting algorithms in actual operations after the October 7 Hamas attacks. However, he revealed that human staff are constantly checking and validating the AI ​​systems’ target recommendations. It was also the human personnel who suggested how to organize the attacks and what weapons to use. “There’s never an algorithm that just works, produces results, and then moves on to the next step,” he said. “At the end of every step involving artificial intelligence, there is human verification.”

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *