Perfect Randomness Realized for the First Time
Source: Phys.org | Published in: Nature — DOI: 10.1038/s41586-026-10521-8
A team of researchers at ETH Zurich has achieved something long thought to be theoretically possible but never before realised in practice: the generation of certifiably perfect random numbers — randomness that can be mathematically proven to be truly random, with no residual predictability whatsoever.
Why Perfect Randomness Matters
Randomness is everywhere in modern technology:
- Cryptography — secure encryption requires random keys that no attacker can predict
- Lotteries and gambling — fairness depends on unpredictability
- Blockchain — consensus mechanisms and validator selection rely on randomness
- Scientific simulations — Monte Carlo methods need high-quality random inputs
- Quantum key distribution — security proofs assume perfect randomness at the source
But most "random" numbers used today are pseudo-random — generated by deterministic algorithms that appear random but are, in principle, predictable if you know the seed. Even hardware random number generators (which use physical processes like thermal noise) cannot prove their outputs are truly random: a sophisticated adversary might be manipulating the physical source.
Perfect randomness closes this gap. It's randomness with a mathematical guarantee.
The Breakthrough: How They Did It
The experiment, led by researchers Renato Renner and Andreas Wallraff (with Anatoly Kulikov as first author), combines three key ingredients:
1. Quantum Entanglement
Two superconducting chips — each containing a qubit — are cooled to near absolute zero and connected by a 30-metre cryogenic tube. Microwave photons are used to create entangled states between the two qubits. Entanglement ensures that the measurement outcomes on the two chips are correlated in a way that cannot be explained by any local hidden variable theory.
2. The Bell Test (Loophole-Free)
The 30-metre separation is critical. It ensures that the two qubits are far enough apart that no information — not even travelling at the speed of light — could pass between them during measurement. This closes the "locality loophole" that has plagued previous Bell tests. Combined with high-fidelity measurements that close the "detection loophole," the experiment provides a loophole-free violation of Bell's inequality — direct proof that the outcomes are not predetermined.
3. The Randomness Amplification Algorithm
Here's the subtle part: even with loophole-free Bell violations, the raw measurement outcomes aren't perfectly random if the measurement settings themselves are chosen using imperfect (pseudo-)randomness. Renner's team solved this with a randomness amplification algorithm that takes weakly random inputs and produces perfectly random outputs.
The process works iteratively: - An imperfect random number generator (seeded by a classical source) chooses the measurement bases - The Bell test is run many times - The amplification algorithm processes the results, distilling the intrinsic quantum randomness and eliminating any contribution from the imperfect seed
The output is provably perfect — certified by the laws of physics, not by any assumption about the hardware.
The "Atomic Clock for Randomness"
The team describes their device as an "atomic clock for randomness" — a physically certified gold standard that other random number generators can be calibrated against. Just as atomic clocks define the second with physical certainty, this device defines randomness with physical certainty.
Unlike atomic clocks, however, the current setup is not practical for everyday use. The cryogenic cooling, superconducting qubits, and 30-metre tube make it a laboratory instrument. But the principles it demonstrates could eventually be miniaturised, much as atomic clocks shrank from room-sized installations to chip-scale devices.
Applications and Implications
| Domain | Impact |
|---|---|
| Cryptography | Perfect keys immune to any computational or physical attack |
| Quantum Security | Enables device-independent quantum key distribution |
| Scientific Computing | Gold-standard randomness for simulations |
| Regulatory | Certified fair lotteries and verifiable random beacons |
| Blockchain | Unbiased validator selection, provably fair protocols |
The work also has deep philosophical implications. It demonstrates that the universe truly is probabilistic at its core — that the randomness observed in quantum mechanics is not a limitation of our knowledge but a fundamental feature of reality.
What's Next
The ETH Zurich team is working on:
- Increasing the bit rate — currently slow, limited by the cryogenic cycle time
- Reducing the physical footprint — smaller cryostats, integrated photonics
- Networked devices — multiple randomness beacons that can cross-certify each other
Commercialisation is likely years away, but the theoretical barrier has been broken. Perfect randomness — once a mathematical abstraction — is now a physical reality.
Bottom Line
Renner, Wallraff, Kulikov, and their team have done something that feels almost philosophical: they have built a machine that produces numbers that are provably unpredictable by any entity in the universe. The achievement is a triumph of quantum engineering and a milestone in the long human quest to understand — and harness — the fundamental randomness at the heart of reality.