The CIA deployed a highly classified tool known as 'Ghost Murmur' to locate and rescue a US airman shot down over Iran last weekend, according to a report. The officer, identified only by his callsign 'Dude 44 Bravo' and described by President Donald Trump as a 'highly-respected Colonel', served as a weapons system officer in an F-15E Strike Eagle that was downed on Friday southwest of Isfahan.
While the pilot was rescued the same day, the second airman was injured, 200 miles behind enemy lines, and armed only with a handgun. He evaded capture for 36 hours in barren wilderness with a bounty on his head. Trained in survival techniques, he activated a Combat Survivor Evader Locator beacon before taking cover in a mountain crevice, but his exact location remained uncertain.
The Ghost Murmur technology uses long-range quantum magnetometry to trace electromagnetic signals produced by a human heartbeat, pairing the data with AI software to isolate it from background noise. Developed by Lockheed Martin's Skunk Works division, it has been tested on Black Hawk helicopters and is intended for future use with F-35 jets.
Speaking at a White House briefing, Trump praised CIA Director John Ratcliffe for a 'phenomenal job' and joked about the classified nature of the tool. Ratcliffe said the agency used 'exquisite technologies that no other intelligence service possesses', comparing the rescue to 'hunting for a single grain of sand in the middle of a desert'.
A source told the New York Post that the technology works like 'hearing a voice in a stadium, except the stadium is a thousand square miles of desert', adding that 'if your heart is beating, we will find you'. The remote Iranian wilderness provided an ideal environment with minimal electromagnetic interference and strong thermal contrast at night.
The source noted that advances in quantum magnetometry, using sensors built around microscopic defects in synthetic diamonds, made it possible to detect heart signals at dramatically greater distances. However, the capability is not omniscient and works best in remote, low-clutter environments with significant processing time.
