A government UAP report is almost engineered to invite the wrong argument. One side scans for the word unresolved and treats every remaining case as an extraterrestrial promissory note. The other scans for balloons and satellites and declares the whole subject solved. Both approaches throw away the part that is actually useful: the machinery of resolution.
AARO's Fiscal Year 2025 report, posted in July 2026, covers events from June 2, 2024 through May 30, 2025 plus older events not previously reported. The office received 319 reports. It resolved 114 of those and another 256 from earlier periods, for 370 resolutions during the reporting period. Every resolved case was attributed to ordinary objects or phenomena, including balloons, satellites, birds, aircraft, unmanned aircraft, a commercial rocket launch, and one manned jet pack.
The real breakthrough was not a craft. It was a model.
AARO says a newly onboarded analytical capability allowed its analysts to resolve 238 reported UAP as satellite flaring. That number matters because it demonstrates something easy to miss in UAP culture: a case can remain mysterious for years not because the underlying phenomenon is exotic, but because the analyst lacks the right reconstruction tool.
Satellite flares are geometry problems. Sunlight reflects from a satellite surface into an observer's line of sight only under specific relative positions. A pilot can experience the result as a bright orb appearing, moving, pulsing, or vanishing against a dark sky. Without ephemerides, observer position, orientation, time, and an appropriate three-dimensional model, the report can look physically strange. Add the geometry and dozens of mysteries can collapse at once.
This is exactly why Cyberdelia treats sensor metadata as part of the sighting. Better analysis does not merely produce better explanations. It changes the historical unresolved count.
Unresolved is increasingly a data-quality category.
AARO transferred 191 of the 319 incoming cases into what it calls the active archive. Those cases lacked enough corroborating data to determine whether the underlying event was natural, conventional technology, or something exceeding known state of the art. Nine reports were considered worth further analysis by intelligence-community and science-and-technology partners.
That structure matters. “Active archive” does not mean “confirmed anomaly.” It means the evidentiary record is too weak for a confident assignment. A blurry clip without range, a narrative without synchronized sensor data, or a report submitted long after the event can remain unresolved forever even if the original object was perfectly ordinary.
The most scientifically valuable UAP program would therefore look boring from the outside. It would standardize timestamps, preserve native sensor products, synchronize platforms, collect weather and traffic context, document optics and processing states, and make sure analysts can reconstruct the geometry later. The prestige object is not the video. It is the dataset around the video.
The maritime case is interesting because it is dense.
The report notes one maritime-domain case submitted by U.S. Navy assets operating off Virginia. The report described approximately 100 airborne UAP and two likely uncrewed surface systems. AARO says it is actively investigating the event with the reporting unit.
That is more useful than a famous isolated light precisely because density creates opportunities for reconstruction. A large event may leave radar tracks, ship logs, environmental observations, communications, electro-optical records, air-traffic data, surface tracks, and independent witnesses. If the files can be synchronized, the event becomes a systems problem rather than a story.
The right question is not whether “100 UAP” sounds dramatic. The right question is how many distinct objects existed, which sensors saw which ones, whether reports refer to the same targets, and what conventional traffic or exercise activity occupied the area at the time.
The nuclear-site numbers belong in a different bucket.
AARO also received 50 reports from nuclear-security and regulatory authorities describing unmanned aerial-system incidents near U.S. nuclear infrastructure, weapons, and launch sites. That was a 177.8 percent increase over the 18 reported the previous year. AARO explicitly says none of those incidents were reported as UAP.
That distinction should be preserved, not blurred for clicks. A drone around a nuclear site can be an immediate security problem without being anomalous in the UAP sense. Conflating the categories makes both investigations worse. The UAS cases demand attribution, counter-UAS capability, perimeter sensing, response authority, and an understanding of adversary reconnaissance. The UAP archive asks a different question: what was observed, and can it be identified from available evidence?
Two interference reports deserve careful handling.
Two reports described electronic or avionic interference that witnesses attributed to nearby UAP. AARO had not determined whether the reported effects were actually caused by those phenomena. This is exactly where causal discipline matters.
An aircraft system can experience interference while something unusual is visible nearby. That establishes temporal coincidence, not mechanism. A serious follow-up needs system logs, frequencies, failure codes, electromagnetic environment, aircraft configuration, and ideally data from unaffected comparison platforms. If a physical coupling exists, it should leave technical fingerprints.
The whistleblower intake numbers are a useful filter.
Between June 2024 and May 2025, 262 people contacted AARO through its authorized reporting mechanism. AARO judged 255 submissions out of scope and identified seven people whose claims warranted follow-up interviews. The important number is not 262. It is the filtering ratio.
Any serious anomalous-claims system needs a triage layer. Claims can be sincere and still fall outside the program's legal scope, lack first-hand knowledge, duplicate other reports, or provide nothing testable. A system that accepts everything as equally probative becomes a rumor warehouse. A system that rejects everything unusual becomes a blind spot. The work is in the filter.
The strongest conclusion is methodological.
The FY2025 report does not provide evidence of captured nonhuman technology. It states that none of the cases AARO resolved indicate advanced foreign breakthrough capabilities, and it says there is no evidence that the U.S. government or a private entity has captured or exploited UAP-derived material. Those are government assessments and should be treated as such, not as metaphysical proof that extraordinary technology cannot exist.
What the report does demonstrate is more concrete: resolution rates change when analytical tools improve; collection bias shapes what gets reported; missing sensor data preserves uncertainty; and high-interest security events can be operationally important without being exotic.
The UAP question gets better when the mystery gets smaller. That is not debunking. It is measurement.
Primary sources
AARO Fiscal Year 2025 Consolidated Annual Report on UAP