What the UAP files actually show

The US Department of War published four tranches of declassified UAP records between May 8 and July 10, 2026. Three hundred and thirty three files. Videos, mission reports, photographs, historical FBI files, CIA program histories, DoE nuclear-facility records, congressional testimony, and one Intelligence Community Assessment.

The volume alone made every UAP outlet blink. The reactions split, roughly, into three camps: this proves everything, this proves nothing, and this is a lot of files.

I don't think any of those three is quite right. I built socialmediaforaliens.com to be a fourth answer.

The gap I noticed

Every existing UAP resource is an inventory. War.gov's own portal is a search-and-download interface. The Black Vault has done heroic FOIA work turning scanned PDFs into a searchable archive. Wikipedia has case pages. Reddit has megathreads.

None of them tell you the one thing you actually want to know when you visit: was there anything real in Release 3?

Not “here are the files.” Not “here's what Reddit thinks.” Not “here's a documentary.” Just a direct answer, from someone who actually read the files, about what each release added, what was recycled from decades of prior FOIA releases, what's overhyped, and what's still missing.

That's an editorial layer, not an archive layer. And so far nobody was doing it.

What I built

Three main features, in the order you'd probably use them.

Per-release verdicts. Four pages, one per PURSUE tranche. Each says what actually shipped that mattered, what was overhyped, what was previously public, and what remains unresolved. Release 04 has forty records, one paragraph tells you which of them changes anything. If you read nothing else on the site, read those four pages.

A top-20 across the whole record, with a specific discipline. Every item is paired: what the source actually says, and what the source does not establish. That second column is the whole point. Most UAP coverage collapses the two. Nimitz Tic Tac establishes that trained aviators observed something with multi-sensor corroboration; it does not establish non-human origin. Gimbal establishes an anomalous infrared return; the prevailing technical explanation (ATFLIR gimbal-lock artifact) accounts for most of it, though it's still contested by other analysts. Hottel establishes that the FBI received a tip; it does not establish the FBI investigated or confirmed anything. Read the twenty items and you'll see what I mean.

Every congressional UAP hearing, embedded. Bray/Moultrie 2022, Grusch 2023, Elizondo/Gallaudet 2024, Restoring Trust 2026. Full videos, sworn witness lists, plain descriptions of what each hearing actually did and didn't establish. See them all.

Under those three surfaces sit eighty two hand-written case pages, a raw catalog of every one of the 333 PURSUE files with inline video and PDF preview, full-text search over the extracted documents, entity pages for named people and places, phenomenon-type browsing, and cross-references throughout.

Every canonical Navy UAP video plays inline. Nimitz Tic Tac. Gimbal. GoFast. USS Omaha. FBI Northeastern Orb collection. Lake Huron F-16 engagement. Not source links to media.defense.gov, actual embedded video that plays in the page.

The 334-page NSA Top Secret / UMBRA UAP FOIA release plays inline as a PDF. So does the AARO FY2025 report, the Robertson Panel report, the CIA U-2/OXCART programs history, Project Sign, the DoE Los Alamos green fireballs conference transcript, the Cometa Report, the Halt memo, the Halt tape, and the Hottel memo.

The idea is you shouldn't have to leave to see the actual evidence.

The editorial move

The pairing discipline is what separates this from every UAP archive I've read. When a case is contested, saying “the object performed anomalous maneuvers” is a claim the source may not actually support. Saying “the aircrews observed what they described as anomalous maneuvers; the sensor data alone does not settle the question” is different. The first collapses the observer's report into an established physical claim. The second preserves the distinction.

That second sentence is harder to write. It requires actually reading the source, being willing to say when a case is genuinely unresolved rather than pushing it into either camp, and being willing to say when a case has a well-supported mundane explanation even if that's the disappointing answer.

The top-20 page enforces this pairing structurally. So does the release verdict format (“what mattered / what was overhyped / what is still missing”). So does the status field on every case (unresolved, explained, contested, administrative). It's not a stylistic choice. It's a discipline that makes the archive useful to someone who's trying to figure out what's real, not just what to believe.

What's honestly not there

I want to name the limits.

I don't have every historical case. Kenneth Arnold's original Army Air Forces signed statement should be on the site and isn't yet. The Belgian F-16 SOBEPS wave report is embedded, but a few other historical primary documents aren't.

Some URLs to government archives (war.gov, aaro.mil, dia.mil, nsa.gov) work in normal browsers but fail for automated tools because of Akamai's TLS fingerprinting. If you visit these directly they load fine, but I can't easily audit them at scale.

Full-text PDF search covers 177 documents across all four PURSUE releases plus the canonical mirrored primaries (Hottel memo, NSA UMBRA, Halt memo, Cometa Report, AARO FY2025, Luna letter, SOBEPS books, Kenneth Arnold's 1952 book). About 15 million characters indexed. What's not yet in the index: photograph-only files (no text to extract) and any post-Release-04 primary docs published after this deploy.

Nimitz Tic Tac and USS Omaha 2019, the two most cited modern cases, don't have primary sensor telemetry in the public record. That's not a limit of the site; it's a limit of what the government has released. The site notes it explicitly rather than pretending we have the raw data.

And I'm not neutral on every question. I default to prevailing scientific explanations when they exist and hold up. When the case genuinely remains unresolved, I say so. When I think a case has been culturally overweighted (Roswell, the Hottel memo), I say that too. If that's a problem for the way you want to read UAP evidence, the site probably won't suit you.

What I hope this becomes

Not another UFO site. A reference layer for people who want to know what the declassified record actually shows, in a form that distinguishes what was released from what it establishes.

If it works, it should be useful whether you're an experiencer, a skeptic, a journalist writing about disclosure, a policy researcher tracking congressional action, or just someone who saw the CBS News segment on Release 04 and wanted to know if the coverage was overstating things.

If you find something wrong, tell me: hello@socialmediaforaliens.com. Errata is public. Corrections are logged.

The site is at socialmediaforaliens.com. Free. No paywall. No email capture.

Start with the four release verdicts or the top twenty items, whichever framing suits how you got here.