Entangled photons could make electronic voting truly anonymous and more secure

Electronic voting promises convenience, speed and easier participation, but it carries a fundamental problem: someone or something normally has to be trusted with the ballots.

Two independent physics experiments have now demonstrated another possibility. Instead of trusting a central server to separate voters from their choices, the systems use quantum entanglement to make that connection inaccessible by design.

Teams at Sorbonne University and the University of Geneva independently tested related quantum voting protocols using entangled photons. Both studies were published in Physical Review Letters, marking important experimental tests of a concept previously developed largely on paper.

The Sorbonne experiment supported a private four-voter, two-candidate election and a second scenario with two pools of four voters choosing among as many as 16 candidates. The Geneva researchers demonstrated a four-person election between two candidates.

“This means that voters don't have to trust each other and voting becomes anonymous to everyone,” said University of Geneva researcher Joey Marcellino.

The core problem resembles an electronic version of dropping anonymous ballots into a hat.

In a physical room, voters can watch the ballot box and gain some confidence that nobody is secretly recording who submitted each ballot. Sending votes to a computer server changes that assumption. Voters must trust the software, the operator and the security of the system storing their information.

The quantum approach attempts to remove that trusted middleman.

Both experiments build on a 2022 protocol designed around Greenberger-Horne-Zeilinger, or GHZ, states. A GHZ state links three or more quantum particles through multipartite entanglement, creating correlations that cannot be reproduced by assigning ordinary predetermined values to each particle.

In the experiments, the qubits were encoded in photons.

Each voter receives one member of an entangled group. When everyone measures their photon in the appropriate way, individual results appear random, but the combined results obey a predictable mathematical relationship.

That shared relationship becomes the temporary secret needed to cast a ballot without revealing who supplied it.

A simplified four-voter example shows how the system works.

Suppose measurements produce four binary values such as 0, 1, 1 and 0. GHZ correlations guarantee that the total number of 1s follows a specified parity, while nobody knows everyone's individual values beforehand.

During a designated voting round, most participants report the value they measured. The person casting the actual ballot can either report that value unchanged or flip it, depending on their choice.

Changing one bit changes the parity of the combined result from even to odd or vice versa. That parity can therefore encode Candidate A or Candidate B.

Observers can determine which candidate received the vote, but the random starting values prevent them from working backward to identify which participant changed a bit.

“Since the exact set of bits we started with was otherwise random, even if we see that someone must have flipped a bit, there's no way to know who it was,” Marcellino said.

Each voter receives a secret position in the sequence, so other participants also do not know when that person is casting the meaningful vote.

Entanglement alone does not solve everything. A malicious operator could theoretically distribute specially prepared particles whose outcomes were already known.

The protocols therefore include verification rounds.

Voters repeatedly test distributed states to determine whether they behave sufficiently like genuine GHZ states. If a source secretly distributes a different state designed to expose votes, those verification tests can reveal the deviation.

The Geneva team confronted another quantum complication: measuring a quantum system generally disturbs it. The same GHZ state cannot simply be fully tested and then reused for voting.

Their solution was probabilistic. For each distributed state, participants randomly decided whether it would be tested or used to cast a vote, with most states devoted to verification. Repeated successful tests increase confidence that the smaller number selected for voting were also valid.

Sorbonne researchers used a related strategy while developing privacy-enhancement methods and support for multiple candidates and separate voting pools.

The Sorbonne team's photon source produced four-party polarization-encoded GHZ states using a layered optical interferometer.

The researchers first demonstrated four voters choosing between two candidates with enhanced privacy. They then simulated a more complicated election involving two pools containing four voters each and as many as 16 candidates.

Handling multiple candidates requires encoding choices in several binary digits. Repeated rounds communicate those bits while preserving the anonymous structure of the underlying protocol.

The photon source could generate GHZ events at roughly 10 hertz, but mechanical components used to adjust polarization during verification reduced the complete protocol rate to around one hertz.

The experiment therefore remains far removed from a national election involving millions of ballots.

The researchers themselves identified another problem with very small voting pools. When only a few people participate, information in the published tally can reveal too much about the group's collective choices even when individual quantum transmissions remain anonymous.

The University of Geneva experiment pursued the same broad goal with an independently modified implementation.

Its researchers generated four-photon GHZ states and successfully carried out the four-party voting procedure. Their experiment was designed to avoid requiring working quantum memories, devices capable of storing delicate quantum states for later use.

That matters because quantum memories remain technologically challenging.

The Geneva group instead discarded photon events that arrived while the system was deciding whether to verify or vote. This reduces efficiency, but it makes a proof-of-principle experiment possible with available equipment.

The researchers now hope to expand the approach through Geneva's developing metropolitan quantum network, potentially placing participants at physically separated network nodes rather than around a single laboratory setup.

Neither experiment shows that quantum voting is ready for public elections.

Large-scale systems would require entanglement distribution to far more users, substantially faster photon sources, dependable optical networks, sophisticated authentication and procedures for dealing with equipment failure and dishonest participants.

The Sorbonne implementation also notes that a complete version of its protocol would benefit from quantum memories. Increasing the number of entangled photons while maintaining high fidelity represents another substantial engineering challenge.

The achievement is therefore not a replacement for conventional voting machines. It is an experimental demonstration that one of electronic voting's hardest trust problems can be approached using the laws of physics themselves.

Eventually, the same anonymous communication primitive could have applications beyond elections. Researchers have suggested anonymous message boards, secret sharing and distributed computing as possible uses.

For voting, however, the idea has unusual significance. Instead of asking citizens to trust that an election administrator cannot discover their choices, quantum mechanics could make that knowledge unavailable in the first place.

These resources explore the theoretical foundations of quantum voting, alternative experimental architectures and the entanglement technologies needed to move secure multiparty protocols beyond the laboratory.

Quantum Protocol for Electronic Voting without Election Authorities: Introduces the 2022 theoretical protocol underlying the new experiments, using multipartite entanglement to provide anonymous, publicly verifiable voting without a trusted election authority. (Physical Review Applied, 2022)

Experimental demonstration of a two-voter quantum anonymous voting prototype with continuous variables: Demonstrates a different experimental approach to quantum voting using continuous-variable quantum technology and fiber links extending to 50 kilometers. (npj Quantum Information, 2026)

Entanglement Routing in Quantum Networks: A Comprehensive Survey: Reviews how entanglement can be routed across practical quantum networks, an essential capability if multiparty voting protocols are eventually distributed among remote users. (IEEE Transactions on Quantum Engineering, 2025)

Research findings are available online in the journal Physical Review Letters.

The original story "Entangled photons could make electronic voting truly anonymous and more secure" is published in The Brighter Side of News.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.