What Are Hash Functions?

Hash functions are deterministic algorithms that map an input of arbitrary length to a fixed-size output, typically called a digest or hash value. The mapping is one-directional: computing the digest from an input is fast, but reconstructing the input from its digest is computationally infeasible. This asymmetry underlies much of modern cryptography and data verification. Hash functions appear across computer science wherever data integrity, authentication, or compact representation is needed.

The intellectual roots of hash functions lie in number theory, combinatorics, and algorithm theory. Non-cryptographic hash functions have existed in computer science since the 1950s, used primarily for efficient data lookup in hash tables. Cryptographic hash functions, which add formal security properties, became central to security engineering in the 1980s and 1990s as digital communications required authenticated data channels.

Security Properties

A cryptographic hash function must satisfy three core security properties. The first is preimage resistance: given a digest value, it must be computationally infeasible to find any input that produces it. The second is second-preimage resistance: given one input, it must be infeasible to find a different input that produces the same digest. The third is collision resistance: it must be infeasible to find any two distinct inputs that yield the same digest. The NIST hash functions project maintains the formal definitions and approved algorithms for federal use, referencing these properties in FIPS 180 and FIPS 202.

Collision resistance is the strongest property and the most difficult to achieve. A function can be preimage-resistant while failing collision resistance. MD5 and SHA-1, once widely deployed, were deprecated after researchers demonstrated practical collision attacks against them. Their failure drove successive rounds of algorithm replacement and new competition processes.

Standardized Algorithms

The Secure Hash Algorithm family, published by NIST, defines the approved digest sizes in use today. SHA-2 (including SHA-256 and SHA-512) remains the most widely deployed standard and underpins TLS, digital signature schemes, and blockchain protocols. SHA-3, standardized in FIPS 202, uses a different internal design called a sponge construction, which offers structural diversity as a hedge against algorithmic weaknesses that might affect SHA-2. The NIST Secure Hash Standard documents both families and specifies their approved use cases in federal information systems.

Variable-output hash functions, called extendable-output functions (XOFs), were introduced with SHA-3. XOFs produce digests of any requested length rather than a fixed size, which is useful in key derivation and post-quantum cryptographic constructions.

Hash Functions in Cryptographic Protocols

Hash functions serve as primitive building blocks in a wide range of security protocols. In digital signature schemes such as RSA-PSS and ECDSA, the signer hashes the message before signing, reducing the signing computation and binding the signature to a compact digest. In HMAC (Hash-based Message Authentication Code), a shared secret is combined with the hash function to produce an authentication tag, providing both integrity and source authentication without a public-key infrastructure. Password storage systems use memory-hard hash functions such as Argon2, which won the Password Hashing Competition in 2015, to slow down brute-force enumeration of credential databases.

Merkle trees, used in certificate transparency logs and distributed ledgers, apply hash functions recursively to build a tamper-evident structure over large sets of data.

Applications

Hash functions have applications across a wide range of engineering and computing disciplines, including:

• Digital signature schemes in public-key infrastructure
• Message authentication codes for network protocol integrity
• Secure password storage and credential management
• Blockchain and distributed ledger data structures
• File integrity verification in software distribution and forensics
• Key derivation in cryptographic protocols such as TLS and SSH