No-Cloning Theorem

Overview

In classical computing, copying information is trivial: you can duplicate a bit perfectly.

In quantum computing, there is a fundamental limitation: you cannot perfectly copy an arbitrary unknown quantum state. This is the no-cloning theorem.

This page states the theorem, gives a clean contradiction proof (based on inner products), and explains practical consequences—without using circuits or gate matrices.

Intuition

A qubit state is not a label you can read out directly. If you try to learn “what the state is” by measuring, you generally disturb it.

Cloning would allow a loophole: if you could copy an unknown state, you could make many copies and measure them in many different bases to reconstruct the state without limitation. Quantum theory does not allow that.

The intuition is not “quantum is mysterious.” The intuition is that quantum states contain continuous information (amplitudes and phases) that cannot be extracted and duplicated perfectly by a physical process that must work for all possible inputs.

Formal Description

Statement (no-cloning theorem). There is no single physical operation that takes an arbitrary unknown state $|\psi\rangle$ and a fixed blank state $|0\rangle$ and outputs two copies $|\psi\rangle\otimes|\psi\rangle$ for every input $|\psi\rangle$.

We will prove this by contradiction using a basic property of valid quantum evolution:

• any allowed evolution preserves inner products between states.

Assume there exists an operation $U$ that clones every state:

$$U\,(|\psi\rangle\otimes|0\rangle) = |\psi\rangle\otimes|\psi\rangle \quad\text{for all }|\psi\rangle.$$

Now pick two (possibly non-orthogonal) states $|\psi\rangle$ and $|\phi\rangle$. Take the inner product of the two input states:

$$\langle\psi|\phi\rangle\,\langle 0|0\rangle = \langle\psi|\phi\rangle.$$

After applying $U$, inner products must be preserved, so this must equal the inner product of the two outputs:

$$\langle\psi|\phi\rangle\cdot\langle\psi|\phi\rangle = (\langle\psi|\phi\rangle)^2.$$

Therefore,

$$\langle\psi|\phi\rangle = (\langle\psi|\phi\rangle)^2.$$

The only complex numbers $x$ satisfying $x=x^2$ are $x=0$ and $x=1$.

So the only pairs of states that could be cloned by the same universal cloner are those with overlap 0 (orthogonal states) or 1 (identical states). Since most pairs of quantum states have overlap strictly between 0 and 1, a universal perfect cloner cannot exist.

This proves the no-cloning theorem.

Worked Example

Consider $|0\rangle$ and $|+\rangle$, where

$$|+\rangle = \tfrac{1}{\sqrt{2}}(|0\rangle+|1\rangle).$$

Their overlap is

$$\langle 0|+\rangle = \tfrac{1}{\sqrt{2}}.$$

If a universal cloner existed, the argument above would require

$$\tfrac{1}{\sqrt{2}} = \left(\tfrac{1}{\sqrt{2}}\right)^2 = \tfrac{1}{2},$$

which is false. Therefore, no operation can clone both $|0\rangle$ and $|+\rangle$ perfectly while also cloning every other possible state.

Turtle Tip

Common Pitfalls

Quick Check

What’s Next

You now have the core multi-qubit concepts: tensor products, entanglement, Bell states, reduced descriptions, and the no-cloning limitation. Next we can start building computational tools on top of these ideas—first by introducing single-qubit transformations more formally, and then by moving toward circuits.