FBI-UAP-D024, FD-302, “Airborne Lights and Triangle,” 2002, 2023-2024
This document is an FBI FD-302, a form the Federal Bureau of Investigation (FBI) uses to record interviews. This FD-302 records a March 2024 interview with a former U.S. military pilot, in which the interviewee describes several incidents potentially involving unidentified anomalous phenomena (UAP) in 2023 and 2024. The interviewee also recalls a 2002 incident potentially involving UAP, while deployed as a military reservist in Afghanistan. An FBI digital rendering based upon the observer’s description accompanies this document, under the title FBI-UAP-D025.
The summary above is sourced from the released file metadata as published to war.gov. The analysis sections below are original to this tracker.
Text of the document
Machine-extracted from the PDF war.gov published (its text layer or OCR), so expect errors, especially in names, numbers and handwriting; the PDF is the authoritative copy. Lines that did not extract as readable text are left out. Showing about 877 words from pages 1, 2, 3 of 3.
PAGE 1
FEDERAL BUREAU OF INVESTIGATION . , was intervi ewe - - · - was referred to the FBI by ~ After being advised of the identity of the interviewing Agents and the nature of the interview, - provided the following information : - stated that starting in October 2023, a l ot of pilots have been seeing the same event occurring, namely seeing lights in the sky. Pilots have been discussing the events over Pil ots typicall y saw them between midnight and 3 AM. - did not know if the events occurred at that time because that was when it was dark enough or if it was because that was when the l i ghts were present. - first observation of the lights occurred in October 2023. - was flying from Boston, MA to Dublin, Ireland where he saw the l ights in the northeast towards Gander, Canada. - was at an elevation of 37,000 feet and believed the lights were hi gher than him, probably at 50,000 feet or higher . The lights were white in color, would start out as dim, and get bri ghter before they eventual ly faded out . The ful l cycle of the lights would range from 10 seconds to one minute. The l ights looked like airpl ane landing lights, but - knew that it was not them, nor did he think it was Investigation on This document contains neither recommendatiollll nor conclusions ofthe FBI. It is the property ofthe FBI and is loaned to your agency; it and its contents are not to be distributed outside your agency.
PAGE 2
Starlink, Starlink flares, satellites, or meteors. It ranged between one to four lights at a time in the sky . Some lights did not move, while others moved in straight lines, someti mes changing direction, and typically did not move in a coordinated fashion. Since spotting the lights the first time, - has seen them approximately ten times . At times the lights would disappear for mi nutes before returning. - has seen the lights for 45 minutes up to two hours . - has seen the lights in all directions except for towards the south. - saw the lights three times in the direction of North Dakota while flying to either Seattle or Phoenix. He has also seen them off the west coast and mid-Atlantic coast. - di d not know how far away they were but ventured the y were between 200-500 miles away. - never saw the lights below the horizon and had not heard about them prior to October 2023 . During a trip on February _ , 2024, - was able to capture video on his Samsung Galaxy S22 phone an event involving white lights. - stated the lights were brighter in-person than on the camera, and the lights were similar in elevation to him or were slightly higher . The video is filmed from the cockpit of his aircraft with the lights turned off. - was located approximately 150 miles northwest of Minneapolis, Minnesota and was flying towards the east. The lights were seen towards the north in the North Dakota region at around 4:36 AM Eastern Standard Time on a clear As captured on the video, white lights were seen, sometimes ten or more present at once, moving around in straight lines and getting bri ghter/dimmer. At the 3:27 mark i n the video, - was not sure if two of the lights were really close together or if it was a single light with a tail. At the 3 : 47 mark, - did not know if it was three separate lights . At the 5:22 mark you can see the flight panel, showing that the airplane was traveling at approximately .80 Mach, and that the airplane was 40- 50 miles northeast of Sioux Fall s and coul d see the KFSD airport on the radar. - never felt unsafe from the white lights, and did not believe the lights were within 100 miles of him . as a reservist , - got activated to go to Afghanistan. One
PAGE 3
night he was outside looking at the stars with another pilot. At approximately 4:30 AM local time, - noticed that the stars were getting blacked out over Bagram. It appeared as if something huge went overhead, from west to east, and was an equilateral triangle shape. The object was moving silently, at a consistent speed and altitude, possibly 150 knots, with no lights visible. - estimated it was 500 feet in size, but he could not fully see the shape of the object. The other pilot with - also observed the triangle, and when it passed over stated, "did you see that?" - stated he did not further discuss the sighting with anyone else nor did he hear anyone else discussing the sighting. Looking through the data, - never found information that could explain what he saw in Afghanistan in not aware of anything that would explain what he had seen. Overall, - believed that not a lot of discussions were happening openly about the events with the lights. - did not think there was as much sti gma surrounding these events, but people were just not used to talking about it.
Where this file fits in the PURSUE archive
This file is one of FBI's 104 files in the Trump PURSUE disclosure. Within the FBI agency block it ranks #96 of 104 by Anomalousness Index. Across the entire 450-file archive it ranks #327 of 450.
That places it in the lower-scoring band of the archive (327 of 450 by score), typical of investigative-record style files where the report is paper-based rather than instrumented.
For the broader cluster context, this file is part of FBI Modern UAP Files - the FBI's modern UAP files: the 2022 Colorado Springs interview and the FBI-authenticated northeastern orb videos. The cluster page walks through all 47 member files with regional grouping, sensor breakdown, and standout analysis.
Anomalousness Index: 55/100
Evidentiary weight that this encounter remains unexplained after conventional analysis. Not a probability of extraterrestrial origin - that number is not honestly computable from the released files and this tracker refuses to publish it.
🤖 AI-ASSISTED SCORING · methodology
The six rubric components break down for this file as follows. Each component has a weighted contribution to the final score; the per-component explanation below describes what this file's particular value on that component means in the rubric's framework.
Reported by a witness with no instrumented record. The lowest tier in the rubric's sensor axis. Eyewitness perception in field conditions, even when the witness is highly credentialed, scores below capture by any instrumented modality. This report reached the federal record through the FBI.
Federal agency personnel (FBI investigators or equivalent) recording the report into the federal investigative system. Investigative credentials, but typically operating in a reactive rather than mission-active posture.
Single-witness or single-instrument capture. This is the corroboration tier for the overwhelming majority of the PURSUE archive on the released metadata - the rubric records the honest limit of the underlying record rather than inferring multi-witness corroboration that the released summaries do not establish. A small number of files with an independent second witness or instrument score on the multi-witness/multi-instrument tier above this one.
No kinematic measurements - speed, acceleration, vector - are published in the released file with sufficient precision to score on the kinematic axis. The rubric does not infer kinematic anomaly from narrative observer estimates. Every file in the archive carries this value, which is itself an observation about the disclosure: kinematic-grade telemetry was not part of what was released.
A conventional candidate explanation has been considered but is not dispositive. Every file in the archive scores this way - reflecting that the underlying release metadata systematically caveats strong determinations in either direction. The released summaries warn against reading them as conclusive analytical judgments, and the rubric respects that.
Catalogued as unresolved with no formal review process having concluded. This is the disposition for a large share of the archive's military infrared captures - the reports are logged into the system as unresolved, but no formal review has finalized. The rubric distinguishes this from 'open after review' because the absence of review is itself a status signal.
Bottom line on the score: the Anomalousness Index of 55/100 reflects evidentiary weight that this specific file's encounter remains structurally unexplained by the rubric's six axes - it is not a claim that the underlying event involved anything non-conventional, and it is not comparable across rubrics that use different weights. For the full per-axis weights and the rubric JSON, see /methodology.
Related files in FBI
Other PURSUE files in the same category, ranked by Anomalousness Index.
FBI-UAP-PR001, “Triangle Orbs,” Northeastern United States, 2021
FBI-UAP-PR002, “Red Orb Rotation,” Northeastern United States, 2022
FBI-UAP-PR003, “Orbs Over the Pond,” 2024
FBI-UAP-PR004, “Northeastern Orb Sighting,” 2025
FBI-UAP-PR005, Digital Recreation, Narrative Statement 3-1, Western United States Event, 2…
FBI-UAP-PR006, Digital Recreation, Narrative Statement 3-2, Western United States Event, 2…
Verification
SHA-256:
This hash is the SHA-256 of the file body war.gov served on the verification date above. War.gov has re-processed some file bodies since first release (re-compression + OCR, no content removed - see /changes); we re-verify and record the change rather than silently serve a stale hash. How to check this yourself →