United States.Colorado Springs 2022 · object over Cheyenne Mountain (FBI/IC, PURSUE)

On February 15, 2022, at 09:35 (Mountain Time), five U.S. Army service members stationed at Fort Carson (Colorado) observed an unidentified airborne object over the Cheyenne Mountain area, on the outskirts of Colorado Springs. The location is not trivial: Cheyenne Mountain houses the hardened underground complex of NORAD / Cheyenne Mountain Space Force Station, one of the country's most sensitive command and early-warning nodes, and Fort Carson is a large adjacent Army base. An anomalous object over that airspace carries, by context, a security reading that a sighting over open country would not.

The episode was documented in the federal system with several pieces that PURSUE declassified in its third tranche (Release 03, June 12, 2026): an FBI FD-302 (the standard format an FBI agent uses to record an interview or statement), an FD-1057 (the Bureau's internal electronic communication on an interview with a forensic artist), a digital rendering of the object reconstructed from the testimonies, and an intelligence-community analysis (ICA-UAP-D001, four pages).

The core of the file is the analytical assessment, and it is worth quoting precisely because it sets the tone. The analysis concluded that the airborne object over Cheyenne Mountain, observed by the five service members at 0935 on February 15, 2022, was 'possibly backscattering of sunlight' (that is, an optical artifact produced by the scattering of sunlight), but the document itself qualifies that conclusion with 'low confidence'. That combination (a plausible prosaic explanation held with low confidence) is exactly what keeps the case in the unresolved category. The analysis itself sums up what the five witnesses said: the object was about 6 miles to the west, over and slightly behind the Cheyenne Mountain silhouette, and was watched for about 30 to 180 seconds; it was 'roughly the size of a large jet and resembling an angular, non-symmetrical potato made of uneven panels', 'completely stationary about 300-500 feet above' the mountain 'while slowly changing shape', with a 'slightly translucent, shimmering white appearance, with defined edges'; and 'despite keeping their eyes on the object, it suddenly disappeared'.

The calibrated reading starts from that self-assessment. The analysis explains what it means by backscattering: with the sun about 27 degrees above the horizon, light reflected off the snow covering Cheyenne Mountain (6 to 12 inches) could illuminate nearby low-level clouds, and a slight shift of the clouds or the sun would make the reflection vanish, which would account for the sudden disappearance. The analyst attributes the low confidence to uncertainty about each witness's field of view, the amount of snow cover and the exact elevation and amount of cloud cover. The witnesses described a clear sky, but the morning weather reports indicated partly to mostly cloudy skies. The document adds that no anomalous data were recorded, that the event did not represent an unknown adversarial capability, and that no aircraft or balloons were active in the area. The dates do not match either: the witness the FBI interviewed in 2025 placed the sighting in early February 2022, around 1 or 2 PM, and the analysis at 09:35 on 15 February. That witness also said that afterwards the five went through what else it could have been, 'such as a blimp or a weather balloon'. In that FD-302 the witness gives further details: they watched it for about 3 to 5 minutes without it changing position relative to the saddle of the mountain, so he did not believe it was drifting in the wind; it was covered in intersecting lines forming 'an abstract polygon pattern'; each of them independently drew what he remembered and the drawings were consistent (a horizontal, white, 'bean-shaped' object); and it vanished at the moment the five 'all simultaneously looked away momentarily' to decide who would fetch a phone from the vehicle, unlike the analysis's summary, in which it vanished while they watched. He added that AARO contacted him in early 2024 and that he responded, but that he 'was never interviewed by AARO or anyone else'. In the FD-1057, the 2024 forensic-sketch interview, a former U.S. Army intelligence officer who was among the five described panels that 'shifted in slow waves' and said the object 'cloaked' in the time it took to turn a head. It is a plausible mundane cause for a daytime sighting with no corroborating instrumentation: the file includes no published radar or telemetry. But the 'low confidence' with which the intelligence community itself holds that hypothesis prevents closing it: five military witnesses, at a maximally sensitive site, saw something the analyst could not explain with certainty. The case's value lies in that documented tension, not in a proven anomaly.

The Colorado Springs file matters for its provenance and for an analysis that states its limits, more than for its phenomenal content. It is a modern, well-chained institutional case: five military witnesses, FBI documentation (FD-302 + FD-1057), a visual reconstruction, and a formal intelligence-community analysis, all over one of the most sensitive command sites in the United States. That chain of custody is precisely what gives a case weight.

But its main lesson is methodological. The analysis itself offers a mundane explanation (sunlight backscattering) and, instead of presenting it as a verdict, flags it with 'low confidence'. That candor is uncommon and exemplary: it acknowledges that a phenomenon can be probably prosaic and, at the same time, remain formally unresolved for lack of data. The case thus calibrates two opposite errors: neither inflating it as evidence of something non-human (it is not), nor closing it as a 'solved case' (the agency itself does not close it). It belongs to the bulk of the phenomenon: sincere sightings, by credible witnesses, over sensitive infrastructure, that the best available assessment leaves at a 'possibly, with low confidence'.