In 1969, Nixon destroyed the entire US bioweapons stockpile over risks AI is now recreating
On November 25, 1969, President Nixon stood at a White House press conference and announced something no other nation had done: the United States would unilaterally destroy its entire offensive biological weapons stockpile and confine all future biological research strictly to defensive purposes. His reasoning, formalized the same day in National Security Decision Memorandum 35, was specific and stark. Biological weapons, he said, have massive, unpredictable, and potentially uncontrollable consequences, and may produce global epidemics and impair the health of future generations. In 2026, researchers testing advanced AI models have reported those same systems demonstrating real ability to assist in designing novel pathogens, with some tests reportedly slipping past intended safety containment.
// Why This Isn’t An Outlier
This document isn’t an isolated case. It’s the same institutional habit the record shows repeating for eighty years.
Build the capability quietly. Use it at a scale the public never approved. Account for it only once a court forces the paperwork into daylight. The Classified Record traces that exact pattern from a 1946 screening memo to the 3.4 million warrantless searches the FBI ran in a single year, using the government’s own documents, not speculation.
Get the book on Amazon →The decision followed a real, extensive internal policy review led by Secretary of Defense Melvin Laird, driven partly by pressure for open congressional hearings on the chemical and biological programs and partly by a National Security Council process that found the Pentagon itself divided on the weapons' military value. The core conclusion that emerged was that biological weapons offered limited, unreliable military utility while carrying a uniquely uncontrollable risk once released, since a pathogen, unlike a conventional weapon, cannot be aimed, recalled, or contained once it begins to spread.
Recent reporting on frontier AI model testing describes systems capable of providing real, substantive assistance in the design of novel biological agents, assistance sophisticated enough that some evaluations reportedly slipped past the safety containment specifically built to catch it. This does not mean an AI system has independently created a working bioweapon. It means the underlying knowledge barrier, once one of the most significant practical obstacles to biological weapons development, is measurably eroding, the exact category of risk Nixon's 1969 decision was built to eliminate entirely rather than manage.
This establishes that a real U.S. president made a formal, documented, unilateral decision to eliminate an entire weapons category specifically because of its uncontrollable and unpredictable consequences, a decision that directly led to the 1972 Biological Weapons Convention. This is confirmed by NSDM 35 itself and the extensive declassified National Security Council record surrounding it.
What this does not establish is that AI-assisted biological design represents an equivalent, government-acknowledged policy crisis today. The comparison drawn here is a historical parallel in the underlying risk logic, not a claim that current AI biosecurity concerns have received the same formal, top-level policy response Nixon's decision represents.
Document: National Security Decision Memorandum 35. The White House, November 25, 1969.
You just read one document. Pro shows you every document that contradicts it.
The Contradiction Engine cross-references any two records in the archive and shows you where the government’s own story changed. The Evidence Chain View maps how one program became the next, sourced at every step. Citations generated correctly, every time. Your own notes, saved against the document.
Start your free month →This site costs real money to run. If it’s useful to you, you can help keep it going →