Skip to main content

Researchers devise a zero-knowledge proof for nuke inspection

Researchers devise a zero-knowledge proof for nuke inspection

Share this story

Nuclear weapons inspection poses a dilemma: how do you train inspectors to identify nukes without revealing top-secret information about how to build them? Spreading detailed nuclear weapons knowledge goes against the principle of non-proliferation, and many countries would object to letting inspectors from a body like the United Nations dig around inside their war technology.

As James Acton of the Carnegie Endowment for International Peace writes:

The problem seems as imponderable as Bilbo’s riddle to Gollum in The Hobbit: "What have I got in my pocket?" A state has a concealed object that it claims is a nuclear weapon. It wishes to provide proof, but the object’s design is largely secret and inspectors are forbidden from making measurements that might reveal classified information.But what if an inspector could verify that a weapon is a nuke without having to look inside?

Let's say the inspectors want to verify that a weapon being destroyed is a genuine warhead and not a "fake" that has been substituted while the real one was stashed somewhere else.

A group of researchers from Microsoft and Princeton University have developed a system for verifying that an item is a nuclear weapon without having to take it apart. It's what's known as a "zero-knowledge proof" — a mathematical concept in which a statement can be proven true without revealing how or why.

The zero-knowledge proof for nuclear inspectors relies on firing neutrons at the warheads in order to produce an image. The inspector doesn't want to see an accurate image of the warhead, because that would reveal too much information. Instead, the inspectors' neutron detectors are preloaded with a negative of the image that would be produced by a real warhead. The inspector will get a null reading if the weapon is genuine. The supplied preloads are shuffled randomly by the inspector, so they'd better match the weapons or foul play will be obvious after a sufficient number of tests.

Implementing such a system is a long way away, geopolitics-wise, but it's an intriguing solution.