Read · 10 min

79 years in 10 minutes

Accessible version of the full analysis. For technical depth, see the per-explanation probabilities and the methodology.

The story

In 3 sentences

  1. 1

    There is a phenomenon institutions have been documenting since 1947, and no single explanation is enough to cover all of it.

  2. 2

    Since May 2026, the U.S. government has been publishing previously classified UAP files in installments, through the PURSUE program. The documents come out with no official conclusion: they neither confirm nor rule out anything.

  3. 3

    How information is being released matters more than what is actually released.

Findings

Five findings

01

Confidence · High

The phenomenon is real

375 cases documented by governments, militaries, agencies and the press between 1947 and 2026, in 75 countries. The reports exist and many are on official paper; some have sensor records (thermal cameras, ground and airborne radar) and others rest only on witnesses. Many have an ordinary explanation; the rest is what this archive tries to measure, case by case.

02

Confidence · High

The reports describe different things

Reports describe very different shapes (discs, spheres, triangles, cylinders, 'tic-tacs', rings) and different modes of interaction with humans, from total indifference to documented physical encounter. A single phenomenon would not produce that variety.

03

Confidence · High

Governments know more than they say

There is a documented chain, signed paper, going from a 1947 memo by General Twining to the Air Force ('the phenomenon is real, not visionary') to the May 2026 presidential disclosure. In between: the Robertson panel labels it a psychological threat (1953), the Bolender memo closes Project Blue Book but admits there are reports 'that do affect national security' (1969), the Wilson-Davis leaks (2002), and David Grusch's sworn testimony to Congress (2023). Every piece is auditable. None is rumor.

04

Confidence · Medium

PURSUE 2026 opens files while new limits on whistleblowers are proposed

The PURSUE release (Presidential Unsealing and Reporting System for UAP Encounters) of May 2026 opens archives to the public with one hand. With the other, a May 26 proposal (from the Office of Personnel Management, the agency that regulates federal employees) introduces a much broader non-disclosure agreement that silences, going forward, anyone who wants to report internal wrongdoing.

05

Confidence · Medium

A 1975 theory that helps read this moment

Astrophysicist Jacques Vallée proposed in 1975 that the intelligence behind the phenomenon works as a control system, one that regulates how much society believes in the topic, like a thermostat. Vallée was talking about the phenomenon, not about governments; but the way PURSUE opens archives, neither fully denying nor fully admitting, is reminiscent of that idea of calibrated openness.

Verifiable facts · 375 cases

What we can actually count.
Probabilities are judgments about these facts.

Evidence quality per case

48

Solid

13% of cases

166

Acceptable

44% of cases

161

Indicative

43% of cases

Solid (S) strong evidence: sensor + multiple witnesses · Acceptable (A) verifiable institutional evidence · Indicative (B) limited evidence: single-witness or without primary verification

Distribution by era

17

Antecedents

‹1946

62

First wave

1946–59

113

Cold War

1960–79

56

Late Cold War

1980–95

47

Before disclosure

1996–16

80

Disclosure

2017–30

Analytical contribution

The taxonomy of disclosure

When an institution faces the UAP phenomenon, the cases show four distinct ways of handling it:

  1. 1Classic cover-up : the official message is 'nothing to see here'
  2. 2Strategic ambiguity : the state shows everything and lets the public decide (this is what PURSUE 2026 does)
  3. 3Pressure ecosystem : politicians, journalists and whistleblowers push together to force openness
  4. 4Leaks : someone decided on their own to release classified information