Can you trust a coin toss? A quantum beacon comes with receipts

Imagine choosing the order of a public list using a random number. Someone dislikes the result and asks a reasonable question: who picked the number, and could they have tried several until they got an outcome they preferred? A statement that the number came from a clever machine does not fully answer the question. A useful public randomness service needs a way to explain where its output came from, when it became available and how others can check the relevant record. That is the appeal of randomness with receipts.

In June 2025, NIST and collaborators described CURBy, a public service combining quantum-generated randomness with a verifiable record of the process. The work connects a foundational physics experiment with the practical problem of making an output inspectable. This is an established research development revisited for its everyday meaning, not a newly announced October 2026 service. The counterintuitive part is that making a random value public can be useful. The equally important qualification is that a public value is not a secret.^[1][2]

Randomness and secrecy are different jobs

A number can be difficult to predict before it is produced and completely visible afterwards. That is often exactly what a public process needs. Everyone can use the same result and inspect the same record without requiring one organiser to keep a private secret. A cryptographic secret key has a different requirement. It must remain secret from an adversary. NIST explicitly warns against using public randomness-beacon values as cryptographic secret keys. A service that publishes its output is not a source of private passwords simply because the output is random.^[3]

This distinction is easy to miss because the word “random” is often used as a general synonym for secure. Security depends on the role a value plays. Public randomness can help coordinate a transparent procedure. Private randomness can help generate secrets. Substituting one for the other changes the system. The useful everyday analogy is a public coin toss before a game. Once the coin lands, everyone is supposed to know the result. The value was useful because it could not be chosen in advance by the person benefiting from it, not because it stayed hidden forever.

Why invoke quantum physics?

The CURBy work uses quantum experiments associated with Bell tests as part of its randomness-generation process. The resulting claims depend on the experimental design and stated assumptions; this is not a declaration that any device carrying a quantum label produces trustworthy output. For a general reader, the key distinction is between merely asserting that a process is random and supplying evidence relevant to that assertion. Physics can contribute to that evidence. It does not remove the need to know how a device was operated, how data were processed or how the result was recorded.^[1][2][4]

A quantum experiment also does not determine the fairness of every process that later uses its output. The random value is an ingredient. The rules that turn it into an outcome remain part of the system and can be badly designed or manipulated. That limitation is not an embarrassment. It is what makes the technology useful to discuss honestly. A better ingredient improves a recipe only when the rest of the recipe is understood.

The organiser can still choose the rules

Consider a hypothetical classroom activity in which a teacher uses a future public random value to decide presentation order. To make the process inspectable, the class would need to agree on the list of participants, the relevant future value and the method that maps that value to an order before the value is known. If the organiser waits to see several values and then chooses a convenient one, the process can be biased even when each individual value is genuinely random. If the list changes after the value is published, the outcome can also change. If the mapping method favours some positions, good input randomness does not repair it.

This example is illustrative, not a claim that CURBy provides a turnkey classroom-allocation application. It shows why a public record is useful only when it is connected to clear rules. The random output and the surrounding commitment procedure solve different parts of the problem. The same principle applies to many transparent selection processes: specify what is fixed before the random value arrives, then make the calculation reproducible afterwards. The details should be designed for the actual application rather than copied casually from a news article.

Receipts change the trust conversation

The NIST account describes a system that links records so that users can examine the provenance of generated values. The point is not simply that a number exists. It is that there is evidence about the process associated with that number. This changes the question from “Do you trust the person who says the draw was random?” to a more specific set of questions. Can the relevant record be checked? Does it correspond to the value used? Were the rules committed in advance? Are the assumptions and limitations clear?^[1]

Not every reader will inspect every technical record. That is true of many systems with public auditability. The value is that inspection is possible and that informed reviewers can test claims. A receipt that nobody is permitted to examine would offer a different kind of assurance. It is also important not to turn a historical service-performance figure into a current guarantee. An early operating interval can demonstrate useful reliability without establishing today's uptime. A public service should be judged on current operational evidence as well as the research that introduced it.

This is not a machine that makes fairness automatic

Fairness can involve eligibility, access, representation, procedure and the consequences of an outcome. Randomness is appropriate for some decisions and inappropriate for others. A transparent random draw does not settle whether the underlying rule was fair in the first place. That is why the most interesting application story is not “quantum physics removes human judgement”. It is “a particular technical tool makes one part of a process easier to examine”. People still have to decide whether that process is suitable and whether its assumptions are acceptable.

The distinction also protects the technology from being asked to solve a problem it was not designed to solve. A randomness service should not be blamed for an unfair eligibility rule, but neither should its physics be used to lend scientific authority to that rule.

A modest benefit with a memorable shape

Public quantum randomness is a useful example of a technology whose practical value may be narrower than its dramatic name—and more interesting because of that. It does not teleport information, replace every source of randomness or make every online system secure. It offers a way to connect an unpredictable output with inspectable evidence. The version worth sharing is therefore not “quantum computers can now make perfect decisions”. It is “a public random value can come with a record that helps people check how it was produced and used”. Keep the second sentence with it: public randomness is not a secret key, and fair use still depends on the rules.

A coin toss is familiar. A coin toss with a reproducible record, a declared future moment and an agreed interpretation is a more demanding idea. Quantum physics can contribute to that picture. The receipts are what make the contribution easier for everyone else to examine.