Two letterheads keep turning up across the declassified record: the Federal Bureau of Investigation and the Central Intelligence Agency. Both documented unidentified anomalous phenomena, on different paperwork and mostly in different places. The archive currently holds 86 FBI files and 19 from the CIA, and reading the two collections side by side says as much about each agency's mission as it does about the sightings.
This article compares what those files actually contain. Every case below is summarized according to what the underlying record states, and each case ID links to its page in the archive, where the document itself can be read. These are government files about reports, investigations, and analyses, not confirmations of any particular explanation.
A First Look at the Numbers
Of the 294 declassified files currently in the archive, the FBI contributes 86 and the CIA 19. That four-to-one ratio reflects how each organization runs into the subject in the first place. The FBI is a domestic law-enforcement agency. It takes reports from the public and from other agencies, and it generates a high volume of standardized paperwork. The CIA is a foreign-intelligence service. Its UAP material tends to show up as a byproduct of collection abroad, which produces fewer documents but often denser ones.
Both agencies are dwarfed in raw count by the Department of War, which holds 143 files and supplies most of the archive's early Cold War history. Still, the FBI and CIA collections are the ones that most clearly show two distinct institutional approaches to the same broad phenomenon, and both can be browsed as standing collections: the FBI UAP files and the CIA UAP files.
How the FBI Documents UAP
The defining feature of the FBI collection is the standardized form. The CIA writes narrative intelligence products. The Bureau memorializes information on the same instruments it uses for any investigation, which makes its UAP files instantly readable to anyone who knows how the FBI works.
Take FBI-UAP-D002, an FD-1057 form holding a first-hand narrative of a UAP encounter near Colorado Springs, Colorado, in 2022. The document is partially redacted, which is typical for witness records carrying personally identifying information, but the FD-1057 structure preserves the essentials: what was observed, where, and the circumstances under which the report was made. The interviewee gave the statement to FBI special agents specifically to help produce a digital artistic interpretation of the incident, and that rendering sits in the archive as its own file, FBI-UAP-D003. The form records the account and stops there.
A second example, FBI-UAP-D009, is an FD-302, the form agents use to write up interviews. This one records a February 2026 interview with a U.S. person describing an orb sighting in the northeastern United States; the sighting itself is dated only 2026. The file's catalog entry adds one more thread: the subject matter described in FBI-UAP-D009 and its sibling interview FBI-UAP-D010, and depicted in the video footage FBI-UAP-PR004, corresponds to reports originating from the same general area in the northeastern United States. That is the full strength of the connection as the record states it, reports from the same part of the country, and both written interviews can be read in the archive.
What the FBI Files Focus On
Read across the 86 files and a few patterns surface:
- Domestic geography. The Bureau's files cluster inside the United States, which lines up with the archive's broader finding that the U.S. is the single most common region overall (82 files).
- First-hand witness testimony. The FBI's value is in raw, contemporaneous accounts collected on the record from named or redacted witnesses.
- Standardized forms over analysis. FD-302 and FD-1057 documents record a report faithfully. They do not adjudicate what it means.
- Recent and historical span. The collection runs from mid-century material into 2026, with cases like FBI-UAP-D009 sitting at the contemporary end.
How the CIA Documents UAP
The CIA collection reads differently. Instead of intake forms, its hallmark is the intelligence information report, the IIR, and the analytical memo. These documents are written to inform decision-makers, and at least one of them preserves something the FBI forms rarely do: a recorded debate about what an object might be.
That one is CIA-UAP-017, a never before released report on a July 2008 UFO sighting at Harare International Airport in Zimbabwe, filed under the title "Placement on High Alert Due to Perceived Aggressive Foreign Posturing." The archive's description of the file comes down to one sentence of substance: individuals debated whether the sighting was an advanced reconnaissance device of a foreign government or of extraterrestrial origins. An ambiguous observation near an international airport, taken seriously enough to generate a CIA report, with the recorded readings running from foreign reconnaissance hardware to an extraterrestrial origin.
A second representative file, CIA-UAP-D001, is a CIA intelligence information report describing human-source collection in the USSR, dated December 20, 1973. The report characterizes its own content as informational, not as finally evaluated intelligence. In section 14, the source describes an incident from the summer of 1973: an airborne, luminous, bright green unidentified object, concentric circles forming around it over a period of several minutes, no sound, and then it dissipated. The source offered no opinion on the nature of the phenomenon and could provide no further details. Where the FBI records a witness on a domestic form, this document reflects collection abroad during the Cold War, UAP material arriving as a downstream product of the agency's core foreign-intelligence mission rather than a dedicated inquiry.
Across the 19 files, the emphasis differs sharply from the Bureau's. The holdings point outward, with documented sightings across the former USSR, Eastern Europe (Hungary, for one), Africa (the Harare case), and Central and South Asia. The formats are intelligence formats, IIRs and analytical memos written to brief decision-makers rather than to document a citizen complaint. And much of the material arrives as a byproduct of HUMINT and other collection rather than as standalone domestic investigations.
FBI vs. CIA: The Core Differences
Set against each other, the two agencies' files map cleanly onto their mandates:
- Volume. FBI 86 files versus CIA 19, reflecting domestic intake versus selective foreign collection.
- Format. The FBI relies on standardized forms (FD-302, FD-1057). The CIA relies on narrative intelligence reports and analytical memos.
- Geography. The FBI looks inward at the United States. The CIA looks outward at events abroad.
- Posture. The FBI records what a witness said. The CIA material runs from raw informational reporting to recorded debate over what an object might have been.
- Sourcing. FBI files originate from public and inter-agency reports. CIA files often originate from intelligence collection operations.
In practice the two collections answer different questions. The FBI file gets you closest to the witness, on the record, near the event. The CIA file preserves what happened to a report once it entered intelligence channels.
Where the Two Agencies Converge
For all their differences, the FBI and CIA files increasingly feed the same modern pipeline, and Colorado Springs is the cleanest example in the archive. The FBI's FD-1057 narrative in FBI-UAP-D002 describes a 2022 encounter there. The archive also holds ICA-UAP-D001, an analysis of the same 2022 incident written by an Intelligence Community partner of AARO, the Department of Defense's All-domain Anomaly Resolution Office. U.S. military service members reported the incident to AARO in 2023. The IC partner assessed, with low confidence, that the reported phenomenon, which observers described as resembling an "angular, non-symmetrical potato," was attributable to sunlight backscattering, sunlight reflecting off mountain snow cover and lighting the underside of low-altitude clouds. That low-confidence assessment contributes to AARO's consideration of the incident, which remains unresolved as of June 2026.
So one incident produced two very different documents: a first-hand FBI field report and a formal analytic product from an AARO partner. Earlier eras handled these reports through ad hoc, agency-specific channels, like the 1949 flying saucer evaluation study in DOW-UAP-D084, prepared at the request of the U.S. Army General Staff's Plans and Operations Division to determine whether the origin could be traced to natural phenomena or the activities of a foreign power. Today, an FBI witness record and an intelligence-community analysis of the same sighting sit in the same declassified release.
Why the Comparison Matters
Read the FBI and CIA holdings together and you learn something neither collection delivers alone: the same phenomenon looks different depending on who is writing it down and why. A Bureau agent filling out an FD-302 and a CIA officer drafting an IIR are answering different questions. The agent asks what this person saw and reported. The intelligence writer asks what the object could be and whether it matters to national security. Holding both documents in view keeps the focus on what the records actually say.
The files that state an assessment state it cautiously. The FD-1057 in the Colorado Springs case is cataloged as an unresolved UAP report. The AARO-partner analysis of the same incident offers a low-confidence explanation, and the case remains unresolved as of June 2026. The CIA's 1973 source offered no opinion at all on the object he described. The records let readers watch two agencies with two different missions describe and assess the same broad question using the tools and language of their own work.
A workable reading order: CIA-UAP-017 for Harare and CIA-UAP-D001 for a Cold War intelligence report, then FBI-UAP-D002 and FBI-UAP-D009 for the shift from intelligence memo to standardized witness form, and finally FBI-UAP-D002 next to ICA-UAP-D001 to see where the two streams now meet.