The first thing you notice on the PURSUE landing page is that it is not a document dump in the usual FOIA sense. It is a structured archive, with case-folder navigation, metadata fields, and a public reporting form that sits alongside the released material. Whoever built it understood that a 12,000-page PDF would have been a worse outcome than a thinner archive that an actual human can navigate.
The second thing you notice is that the contents are extraordinarily uneven.
The Apollo material
The headline-grabbing pieces are two clusters of photographs and transmission transcripts from Apollo 12 (November 1969) and Apollo 17 (December 1972). Most of the underlying frames have appeared somewhere — NASA archives, Lunar and Planetary Institute scans, third-party photographic compilations. What is genuinely new is metadata: scan-quality flagging, cross-references to mission air-to-ground transcripts, and in two cases, accompanying staff annotations that appear to have been redacted from earlier public versions.
The honest read is that the Apollo material is interesting context but not a smoking gun. Anyone looking for a clear ET craft on the lunar horizon will not find it here. What the material does usefully do is establish a pattern: there are official records that have been quietly held back from public view not because they prove anything sensational, but because the review process simply never ran. PURSUE has begun running it.
The case folders
Far more substantive — and far more frustrating — are what the portal calls “Resolved-Status: Anomalous” case folders. These are AARO incidents where the office's own analytic process couldn't conclude the encounter was a known platform, balloon, sensor artifact, or natural phenomenon.
Roughly 60 such folders have been published in the initial release. The folders generally include:
- Date, location, and reporting unit (some redacted)
- A short narrative summary
- Sensor data — radar tracks, infrared signatures, occasional video stills
- An analytic conclusion of “unresolved” with reasoning
What's missing — and this is the structural frustration — is the raw sensor feed. AARO's narrative summaries describe radar tracks behaving in non-aerodynamic ways. The portal does not, in this drop, give analysts the underlying signal data needed to verify that interpretation independently. That's not a failure of the release per se; it's a real classification problem (sensor capabilities reveal sensor limits to adversaries). But it means the case folders are something closer to government testimony about what was observed than raw evidence of what was observed.
“The folders are government testimony about what was observed — not raw evidence of what was observed.”
The FBI memos
A third tranche surfaces FBI memoranda referenced in earlier FOIA litigation but never fully released. These are mostly historical — late 1940s through the 1970s — and overlap substantially with material previously surfaced through The Black Vault and other long-running disclosure repositories. Researchers cataloging the new material against existing FOIA archives will find the diff is small but non-zero.
The intake form
One often-overlooked element of PURSUE is the public intake form. Active-duty service members, pilots, and air traffic controllers can now report directly through the portal — not just through AARO's prior channels. The form includes whistleblower-protection language and a clear chain-of-custody pathway to AARO's analytic queue.
Whether that intake produces a meaningful uptick in reports is one of the metrics worth watching over the next 90 days. AARO's existing caseload already exceeds 2,400 reports. A surge of new intake without a corresponding surge in analytic capacity is the operational risk to track.
What the architecture tells us
If you treat PURSUE as a systems object rather than a political event, two things stand out.
First, this is built to scale. The case-folder schema, the metadata fields, the structured intake — all of it suggests an architecture designed for ongoing rolling release rather than one big drop. That tracks with the administration's stated plan to release in tranches over the next year.
Second, the redaction policy is conservative. Several published case folders include “[redacted]” on what appear to be sensor-platform identifiers and reporting-unit designations. That's expected. What is more notable is that some folders also redact narrative-level content — descriptions of the encounter itself — which is harder to justify on classification grounds and will draw scrutiny.
What to read it against
Three reference points for serious readers:
- AARO's FY24 Consolidated Annual Report on UAP for baseline analytic methodology.
- The 2023 House Oversight subcommittee hearing transcripts, which set the floor for what whistleblowers have alleged but not documented.
- The conferenced FY2026 NDAA UAP provisions, which define what AARO is now required to brief Congress on going forward.
PURSUE doesn't answer the question disclosure advocates have been asking for two decades. It does, for the first time, give the question a public file system. That's not nothing.