Homomorphic Encryption
Pronunciation: HOH-moh-MOR-fik en-KRIP-shun
Also known as: Encrypted computation, Privacy-preserving computation on ciphertext
Definition
Homomorphic encryption is a cryptographic technique that allows selected computations to be performed on encrypted data without first exposing the underlying plaintext to the computing party. It differs from ordinary encryption at rest because ordinary encryption generally requires decryption before processing, while homomorphic schemes preserve a mathematically meaningful relationship through computation.
Overview
Meaning and scope
Homomorphic encryption is a cryptographic technique that allows selected computations to be performed on encrypted data without first exposing the underlying plaintext to the computing party. It differs from ordinary encryption at rest because ordinary encryption generally requires decryption before processing, while homomorphic schemes preserve a mathematically meaningful relationship through computation.
Mechanism and operation
Operational implementation should select a scheme and security level appropriate to the calculation, protect keys, validate implementations, measure accuracy and performance, minimize plaintext exposure, and document residual leakage. The principal failure modes include incorrect parameter selection, excessive computational cost, implementation flaws, side channels, key compromise, output leakage, and claims of privacy that exceed the actual threat model. Useful measures include encrypted workload coverage, latency overhead, key rotation status, failed computation rate, and privacy tests against defined attackers. Operationally, teams should select a scheme and security level appropriate to the calculation, protect keys, validate implementations, and measure accuracy and performance.
Practical payment example
A research team tests computing an aggregate statistic over encrypted payment records. It measures performance and leakage boundaries before suggesting the method for a production fraud-screening workflow.
Operational checks
Assess the exact scheme, supported computation types, key custody and performance overhead. Encrypted aggregate processing does not automatically hide access patterns or all intermediate metadata. A payment platform should test result correctness, failure cases and compatibility with its fraud-detection requirements before deployment.
Distinctions and limits
Key risks include incorrect parameter selection, excessive computational cost, implementation flaws, and side channels. Review of Homomorphic Encryption should determine whether safeguards addressing a cryptographic technique that allows selected changed exposure in practice, not merely whether a document or setting existed.
Related reading
Related definitions: Privacy-Preserving Analytics, Re-identification Risk.
Example
A research team tests computing an aggregate statistic over encrypted payment records. It measures performance and leakage boundaries before suggesting the method for a production fraud-screening workflow.
Comparison
Homomorphic Encryption is closely connected to Privacy-Preserving Analytics, Re-identification Risk, and Privacy Risk Assessment. It differs from ordinary encryption at rest because ordinary encryption generally requires decryption before processing, while homomorphic schemes preserve a mathematically meaningful relationship through computation.
Key Takeaway
Homomorphic encryption is a cryptographic technique that allows selected computations to be performed on encrypted data without first exposing the underlying plaintext to the computing party.
Sources
- NIST Workshop on Privacy-Enhancing Cryptography 2024 — NIST (2026-08-03)
- NIST Privacy Framework — NIST (2026-08-03)
- Recommendation for Key Management, NIST SP 800-57 Part 1 Rev. 5 — NIST (2026-08-03)