What Is PURSUE?

PURSUE - the Presidential Unsealings and Reporting System for UAP Encounters - is the Trump administration's program for declassifying and publicly releasing the U.S. government's accumulated Unidentified Anomalous Phenomena records. The first release went live at war.gov/UFO on May 8, 2026. Here is how the program is structured, what was in Drop 01, and what the disclosure does and does not establish.

The program in one paragraph

PURSUE is structured as a rolling-release declassification program. Rather than dumping a single archive at one point in time, the program publishes UAP records in periodic "drops" hosted on the U.S. Department of War's public website at war.gov/UFO. Each drop adds files to the canonical CSV catalogue that war.gov maintains, accompanied by released metadata describing each file - file type, the originating agency, mission or case identifiers, brief content summaries, and where applicable, sensor and witness context. Files are released into the public domain under 17 U.S.C. § 105 (works of the U.S. Government), so anyone can mirror, quote, or analyze them without licensing constraints.

Drop 01 by the numbers

The first PURSUE release, published on May 8, 2026, contained 161 distinct file URLs distributed across four federal agencies. The release covered records from 1947 through 2026 - nearly eight decades of accumulated federal documentation of UAP encounters and investigations. The breakdown:

By file type, the release breaks down into 119 PDFs (investigative records, mission reports, transcripts, technical proposals), 28 videos (mostly infrared captures from U.S. military mission platforms), and 14 image files (lunar surface photography from NASA, plus FBI archival photographs). The full breakdown is searchable on this tracker; the top-10 page shows the highest-scoring files on this site's evidentiary rubric, and the Pentagon, FBI, NASA, and State Department category pages list every file in each agency.

What the agencies actually contributed

Each agency's PURSUE submission is structured around an internal organizing logic. Understanding what each contributed clarifies why the rubric scores them differently:

Department of War (82 files)

The largest single contribution, dominated by short infrared video captures from U.S. military mission platforms submitted to the All-domain Anomaly Resolution Office (AARO). Most are CENTCOM submissions from 2020-2026, with a smaller Indo-Pacom contingent. The DoD release also includes accompanying mission reports (DoW-UAP-D and DoW-UAP-VM serial conventions) that contextualize each video. The 27 files at score 66 - the densest tier in the archive - are nearly all from this block.

Federal Bureau of Investigation (57 files)

The FBI's contribution centers on its internal case file 62-HQ-83894, which aggregated UFO and "flying disc" investigations between June 1947 and July 1968. The case file is broken into 10 sections (the chronological investigative record), 7 serial-numbered records (specific catalogued items), and 1 sub-file. The PURSUE release includes pages that the public FBI Vault has not previously posted. The remaining FBI files are individual records adjacent to the central case. See our deep dive on 62-HQ-83894 for the file-by-file walkthrough.

National Aeronautics and Space Administration (15 files)

NASA contributed a mix of Apollo-era lunar photography (Apollo 12 and 17 missions), Gemini-era astronaut audio (notably the December 1965 Borman/Lovell Gemini 7 file, which is the highest-scoring single file in the archive), and Skylab crew debriefings. NASA's photographic submissions are notable because NASA modified the originals before release to highlight specific regions of interest containing "unidentified phenomena" - the agency did not assert what was in those regions, only that they were flagged. See our Apollo 12 deep dive.

Department of State (7 files)

The smallest contribution: diplomatic cables forwarded back to the State Department from U.S. embassies reporting foreign-government or foreign-aviation-authority UAP communications, spanning 1952-2025. The releases include cables from Papua New Guinea, Tajikistan, Kazakhstan, Turkmenistan, Georgia, and Mexico. These files document what foreign governments told U.S. diplomats about UAP encounters - the State contribution is a record of diplomatic reporting, not a record of U.S. domestic observations.

The May 11 CSV revision

On May 11, 2026 - three days after the initial drop - war.gov restructured the canonical PURSUE CSV without public announcement. The file count appeared to change: the CSV went from 161 rows to 158 rows. This tracker's auto-poller caught the change within hours and published a verified diff at /changes.

The actual change, verified by URL-set comparison against the May 8 snapshot:

• Zero PDFs were added, zero PDFs were removed.
• Zero videos were added, zero videos were removed.
• Nine PDFs that previously had a single CSV row now have 2-4 CSV rows each (12 extra rows total) - the same PDF cross-referenced from multiple incident-report rows.
• One PDF had its storage slug renamed (the file content unchanged - byte-identical SHA-256, verified).

The 161 → 158 row delta is fully explained by those two facts. No files left the disclosure, and the row reduction is an artifact of how war.gov chose to represent multi-incident reports in the CSV. This is the kind of structural change that surface-level reporting easily mistakes for "files removed" or "redactions added," which is why this tracker maintains the URL-set diff as the authoritative comparison rather than relying on row counts.

What is NOT in PURSUE

The boundaries of what PURSUE actually establishes are worth stating explicitly, because the disclosure has been described in some coverage in terms that overstate it:

• PURSUE does not establish that any of the released encounters involved extraterrestrial craft. Each file's released metadata explicitly caveats that the descriptions provided should not be interpreted as analytical judgments about the nature of the observation.
• PURSUE does not contain kinematic-grade telemetry. The released videos show area-of-contrast tracking with timestamps but not published speed, acceleration, or trajectory measurements at analytical precision. Observer-estimated speeds appear in some accompanying mission reports but are characterized as estimates.
• PURSUE does not exhaust the federal UAP record. The release is what the program has chosen to publish so far. The rolling-release structure implies more is forthcoming; the file count almost certainly grows in subsequent drops.
• PURSUE does not include the AARO formal-review conclusions for most files. The 27-file score-66 cluster, for example, is catalogued as "unresolved with no formal review" - meaning AARO has the report logged but the office's review process has not concluded for those reports.

How this tracker reads PURSUE

This site is an independent, non-governmental project. It exists because the official war.gov interface is a flat list of file thumbnails with no full-text search, no transcripts on the videos, no per-file analysis, and no cross-file comparison. The tracker addresses each of those gaps:

• Full-text search across all PDFs and video transcripts at /search.
• Whisper-generated transcripts on every video file. The Borman/Gemini 7 audio is the highest-stakes example - the original release is just an audio file, and the transcript is what makes the astronaut's words searchable and citable.
• An open six-axis evidentiary rubric applied to every file, scoring on sensor quality, witness credibility, corroboration, kinematic anomaly, mundane-explanation availability, and official disposition. The rubric and weights are published as JSON at /data/scoring-rubric.json so anyone can recompute every score. The rubric measures evidentiary weight that an encounter remains structurally unexplained after conventional analysis - it explicitly does NOT publish a probability of extraterrestrial origin, because that number is not honestly computable from the available data and this tracker refuses to invent one.
• SHA-256 verification on every file against the war.gov original, so the mirror is provably the same bytes the government released.
• An automated poller running on a public GitHub Action that checks war.gov every 30 minutes during U.S. weekday business hours (hourly otherwise) and opens an issue on this repo when the CSV changes. This is how the May 11 revision was caught within hours, and how Drop 02 will be caught when it lands.

Where the program goes from here

Drop 02 has not yet been announced with a firm date. The original disclosure announcement implied a regular cadence, and tracking signals point to a release in the early summer 2026 window. This site's auto-poller is the canonical detector - the moment war.gov pushes a new CSV, the poller catches it and the site re-indexes automatically with the new files added to the manifest, scored against the rubric, and (for videos) transcripted. The verified diff against Drop 01 will appear at /changes within hours of the release.

Whether or not PURSUE eventually establishes anything definitive about the nature of UAP, the program has already done something concrete: it has moved a substantial body of federal records from internal classification into the public domain. The files are now citable, searchable, and verifiable byte-for-byte against the government source. That is the program's actual product so far, and what this tracker indexes.