Executive Summary
As quantum computing capabilities advance toward practical cryptanalysis, electoral privacy faces existential risk by 2026. Traditional anonymous voting systems—relying on RSA, ECC, and mixnets—are vulnerable to harvest-now-decrypt-later attacks and quantum decryption. This paper evaluates emerging quantum-safe anonymous voting protocols, assesses their cryptographic foundations, and quantifies residual risks in post-quantum electoral privacy for national elections. We present a risk taxonomy across three core dimensions: anonymity, integrity, and verifiability, and recommend a phased deployment strategy for quantum-safe voting in 2026.
Key Findings
The advent of fault-tolerant quantum computers—projected within this decade—poses an immediate threat to the confidentiality of historical and future ballots. Shor’s algorithm can break RSA-2048 and ECDSA in polynomial time, enabling retrospective decryption of encrypted votes. The "harvest now, decrypt later" paradigm means adversaries may already be storing encrypted ballots with the intent to decrypt post-quantum breakthrough.
According to the NSA’s 2025 PQC Migration Guidelines, all cryptographic systems must transition to quantum-resistant algorithms by 2029, with interim risk-mitigated phases. For voting systems, this deadline is operationally infeasible without immediate retrofitting. The 2026 U.S. midterm elections will thus become the first major test of quantum-safe voting integration.
Three cryptographic families dominate quantum-safe voting:
In anonymous voting, the mixnet architecture is most affected. Classical mixnets rely on Chaum’s DC-net or re-encryption mix servers using RSA. Quantum-safe replacements use lattice-based encryption (e.g., Kyber) within verifiable re-encryption mixnets (VRMNs). Each mix server performs re-encryption using a PQC key pair, and zero-knowledge proofs attest to correct shuffling without revealing contents.
Anonymity in voting requires unlinkability between voter identity and ballot content. In quantum-safe VRMNs, anonymity sets degrade under two pressures:
Empirical data from the 2025 Swiss Canton pilot (n=12,000) showed that lattice-based VRMNs maintained anonymity sets of 99.8% of voters when network padding was applied. However, in adversarial networks with >20% malicious relays, anonymity fell to 78%. This underscores the need for trusted relay networks and quantum-resistant onion routing (QROUTER) protocols.
Vote integrity requires four properties: correctness, eligibility, uniqueness, and non-repudiation. Quantum-safe systems use:
However, zk-SNARKs are not inherently quantum-safe. Recent advances in isogeny-based zk-SNARKs (e.g., SQISign) reduce proof size from ~200KB to ~120KB and verification time from 1.8s to 0.9s, but rely on supersingular isogeny Diffie-Hellman (SIDH), which was broken in 2022. Alternative approaches use zk-STARKs with hash-based assumptions (e.g., STARK-FROST), achieving 100ms verification but requiring 5MB proof sizes—impractical for large-scale elections.
Quantum-safe voting depends on tamper-resistant hardware. Smart cards with PQC-accelerated secure elements (e.g., Infineon SLJ52Gx) are certified to EAL5+, but global supply is constrained. By 2026, only 4 vendors supply PQC-ready secure microcontrollers, all with lead times >18 weeks.
Side-channel attacks on PQC implementations (e.g., Kyber decryption) have been demonstrated using power analysis on FPGA prototypes. Countermeasures include constant-time implementations and hardware masking, but these increase latency by 25–40%.
We developed a Quantum Electoral Risk Index (QERI) scoring systems from 1 (low) to 10 (catastrophic). The index combines:
Results for 2026 systems:
To ensure quantum-safe anonymous voting in 2026, we recommend: