Homomorphic encryption (HE) is a type of encryption that allows certain calculations to be carried out on data while it remains encrypted. The party doing the work never sees the underlying values; when the owner decrypts the output, it matches the result of performing the same calculation on the plain data. Partially homomorphic schemes support only one operation, such as addition in the Paillier scheme or multiplication in textbook RSA, while somewhat homomorphic schemes allow a limited number of both.
Fully homomorphic encryption (FHE), which permits arbitrary computation, was first shown to be possible by Craig Gentry in 2009, though it remains far slower than ordinary processing. In finance, HE is studied for sharing data between institutions for fraud and AML analysis without exposing customer records, and in blockchain for confidential smart contracts. It is one of several privacy-enhancing technologies, alongside multi-party computation and zero-knowledge proofs.
Ref 1300000313