Afghanistan.Afghanistan · the edited video of a “sphere” among the clouds

According to the title that comes with it, the video was recorded on November 23, 2020 over Afghanistan. It runs 47 seconds and shows an area of contrast crossing the frame among clouds. The Pentagon office that investigates these phenomena, AARO (the All-domain Anomaly Resolution Office), assesses as "likely," without confirming it, that it comes from an infrared sensor aboard a U.S. military platform operating in November 2020 in the area of Central Command (CENTCOM), the U.S. military command for the Middle East and Central Asia. Its record identifies no coordinates, platform, unit or crew.

In June 2024, someone uploaded the file to a classified network under the title "Spherical UAP over AFG in and out of clouds 23 Nov 2020." According to AARO, that title was given by whoever uploaded the file and is not a Department of War designation. AARO adds that "this media was digitally altered prior to its upload to a classified network, and is presented as received."

The clip is a three-part edit separated by black screens with text cards. After a first card announcing a zoom ("zoomed in"), an area of contrast appears near the top left corner and crosses from left to right until it blends into the background. A second card announces sharpening, zoom, motion tracking, contrast enhancement and slow motion at 60 percent ("sharpened, zoomed, motion tracked, contrast enhanced, slow to 60% speed"), and the scene replays that way. The last part, labeled "original video," shows an area of contrast near the top center of the frame that moves down to the left, then to the right, and is lost against the background. AARO describes only "an area of contrast," not a spherical shape.

The video was already circulating before the government released it. On June 17, 2025, Jeremy Corbell released it on his Weaponized podcast with a date (November 23, 2020, 11:13 UTC) and coordinates in eastern Afghanistan that AARO's record does not include, and a thread analyzing it opened on the Metabunk forum that same day. Corbell and George Knapp, who said they had spent about two years authenticating the material (Knapp called the source "legitimate"), presented it as footage from a high-altitude Air Force intelligence and reconnaissance platform over the Afghanistan-Pakistan border: a disc 200 to 400 meters across moving in and out of the clouds with no plume or heat signature. NBC News ran it as a "disc-shaped" UAP.

On March 6, 2026, eight members of the House of Representatives requested access to 51 records potentially related to these phenomena. In responding, AARO located a batch of 50 videos on a classified network, this one among them, cataloged as DOW-UAP-PR055, and warned, without saying which, that "many of these materials lack a substantiated chain-of-custody": there is no way to trace who recorded them or whose hands they passed through. On May 22, 2026 it was released under a public domain license on DVIDS, the Department's imagery distribution service, as part of the second release of the PURSUE program. None of Corbell and Knapp's details (platform, border, shape or size) appears in AARO's record.

The title that names the video was not written by the Department of War but by whoever uploaded it to the classified network in 2024, and the file is a version edited by an unknown person, not the original sensor recording. AARO does not describe a sphere or a disc but "an area of contrast"; the shape, size and platform now in circulation come from the people who released it in 2025, not from the official record.

The ordinary readings proposed in the Metabunk thread are of two kinds: an artifact of the camera itself (a water drop on the sensor window, or sun glare in the lens matching the color of the cloud) or a real object, such as a drifting balloon; the thread reached no agreement. Mick West, who since June 2025 has leaned toward "an internal reflection of the sun" in the optics, admitted in May 2026 that he did not recall an example of that kind of camera having an internal reflection, and left the case as lacking information.

Weighing any of these readings would require the estimated speed, the altitude of the object and of the cloud layer, and the type of platform that was recording, which are the data AARO cross-checks in its case resolution reports to tell, for example, a balloon from a bird in infrared footage. None of them is available here, and the editing adds a prior step: any speed or trajectory analysis would first have to establish which parts are the original recording and which are the enhanced slow-motion version.