Narrative Goldmine

encryption

9 min read

Properties

Type
  • Class
Resource
  • urn:ngm:class:encryption
Domain
  • security
Status
  • stable
Maturity
  • mature
Quality
  • 0.74
Is a
Requires
Enables
Generated
  • by: process:vault-migrate/1.0 · at: 2026-09-22T12:41:39.024503178Z
Bridges to
Contrasts with
Implements
Label
  • Encryption
Links
    Page resource
    • urn:visionflow:page:bf86075ceab2af34c4bf04695936f6522b383cee6b0d377a410b470499ab5485
    Related to
    Standardized by
    Supports
    Uses

    Encryption is the cryptographic process of transforming plaintext data into ciphertext using a defined algorithm and secret key, rendering the data unintelligible to any party that does not possess the corresponding decryption key, thereby ensuring confidentiality. Symmetric schemes such as AES-GCM use a single shared secret for both encryption and decryption, providing authenticated encryption with associated data (AEAD) in a single pass; asymmetric schemes such as RSA-OAEP and ECDH use mathematically linked key pairs where the public key encrypts and the private key decrypts. Hybrid constructions combine both paradigms — using asymmetric key exchange to establish a shared session key and then symmetric ciphers for bulk data — as exemplified by TLS 1.3. Encryption is foundational to data-at-rest protection, data-in-transit security, end-to-end encrypted messaging, confidential computing, and post-quantum cryptography.

    Overview

    • Encryption predates the digital age: classical ciphers such as the Caesar cipher and Vigenère cipher demonstrated the principle of transforming readable text into unintelligible form. Modern encryption is grounded in computational hardness assumptions — for symmetric ciphers this is the infeasibility of exhaustive key search; for asymmetric ciphers this is the intractability of mathematical problems such as integer factorisation (RSA) and the elliptic curve discrete logarithm problem (Elliptic Curve Cryptography).
    • Encryption operates at multiple layers of the computing stack:
      • Data at rest: files, database records, and storage volumes are encrypted to protect against physical theft or insider access.
      • Data in transit: network protocols such as Transport Layer Security and IPsec encrypt data as it traverses networks, protecting against eavesdropping and man-in-the-middle attacks.
      • Data in use: Confidential Computing extends encryption into the execution domain using hardware Trusted Execution Environments (TEEs), keeping data encrypted even during processing.
      • End-to-end encryption (E2EE): platforms such as Signal Protocol and WhatsApp implement E2EE so that service operators cannot access message content — only communicating endpoints hold keys.
    • The security of an encryption scheme depends on three factors: the strength of the underlying algorithm, the length and quality of the key, and the secrecy of the key material. Key Management is therefore as critical to security as algorithm selection.

    Key Mechanisms

    • Symmetric Encryption
      • Uses a single shared secret key for both encryption and decryption.
      • The dominant algorithm is AES (NIST FIPS 197), operating in modes such as GCM (Galois/Counter Mode), which provides Authenticated Encryption with Associated Data (AEAD), combining confidentiality and integrity in a single operation.
      • ChaCha20-Poly1305 is a widely deployed alternative, particularly in software implementations on devices lacking AES hardware acceleration.
      • Key lengths: AES-128 (128-bit key), AES-256 (256-bit key); 256-bit is preferred for long-term security against quantum attacks.
    • Asymmetric Encryption
      • Uses a public key for encryption and a mathematically linked private key for decryption.
      • RSA-OAEP is the canonical asymmetric encryption scheme; Elliptic Curve Cryptography underpins ECDH (key exchange) and ECIES (encryption).
      • Asymmetric operations are computationally expensive and are generally used only to encrypt small payloads (e.g., a symmetric session key) rather than bulk data.
    • Hybrid Encryption
    • Authenticated Encryption
      • AEAD modes (AES-GCM, ChaCha20-Poly1305) provide simultaneous confidentiality and integrity, preventing ciphertext manipulation attacks.
      • Without authentication, ciphertext malleability can allow attackers to modify messages without detection — a critical flaw exploited in padding oracle attacks.
    • Post-Quantum Cryptography
      • Quantum Computing threatens asymmetric schemes based on factorisation and discrete logarithms; Post-Quantum Cryptography algorithms rely on problems believed hard for quantum computers, such as module lattice problems.
      • NIST FIPS 203 standardises CRYSTALS-Kyber (ML-KEM) for key encapsulation; FIPS 204 standardises CRYSTALS-Dilithium (ML-DSA) for digital signatures.
      • Symmetric encryption (AES-256) is considered quantum-resistant with doubled security margin due to Grover’s algorithm, requiring only a key-length increase rather than algorithm replacement.
    • Homomorphic Encryption
      • Homomorphic Encryption allows computation on ciphertext without decryption, producing encrypted results that when decrypted match the result of operations on the plaintext.
      • Fully homomorphic encryption (FHE) supports arbitrary computation; partially homomorphic schemes (e.g., Paillier, ElGamal) support a restricted operation set.
      • Active research area with applications in Federated Learning, privacy-preserving analytics, and Secure Multiparty Computation.

    Applications and Use Cases

    • Web and network security: Transport Layer Security encrypts virtually all modern web traffic; HTTPS, SMTPS, IMAPS, and FTPS are TLS-wrapped variants of plaintext protocols.
    • Messaging and communications: Signal Protocol implements the Double Ratchet Algorithm for forward secrecy and break-in recovery; adopted by Signal, WhatsApp, and others. Matrix/Element implements End-to-End Encryption via Megolm for group messaging.
    • Storage encryption: Full-disk encryption (BitLocker, LUKS, FileVault) protects laptops and servers against physical theft. Database-level encryption (Transparent Data Encryption in SQL Server, Oracle) and object-level encryption in cloud storage (AWS S3 SSE, Azure Storage Service Encryption) protect data at rest.
    • Blockchain and cryptocurrency: Blockchain systems rely on asymmetric cryptography for wallet addresses and transaction signatures; Elliptic Curve Cryptography (secp256k1 for Bitcoin, alt curves for Ethereum) underpins ownership proofs. Encrypted mempools and commit-reveal schemes protect transaction privacy.
    • Confidential computing: Trusted Execution Environments such as Intel TDX and AMD SEV-SNP extend encryption to data-in-use, enabling Confidential Computing for cloud workloads and AI model inference without exposing data to cloud providers.
    • Privacy-enhancing technologies: Zero Knowledge proofs can prove knowledge of a plaintext without revealing it; combined with encryption, they enable verifiable computation on private data.
    • Federated learning and AI: Federated Learning deployments use encryption and Secure Multiparty Computation to aggregate model updates without exposing individual client gradients, addressing privacy concerns in distributed AI training.
    • Regulatory compliance: GDPR, HIPAA, PCI-DSS, and other regulations mandate encryption of personal and financial data, driving widespread enterprise adoption of encryption at rest and in transit.
    • VPN and remote access: IPsec and WireGuard encrypt network tunnels for virtual private networks, enabling secure remote access and site-to-site connectivity.

    Standards and Governance

    • NIST FIPS 197 — Advanced Encryption Standard (AES); mandatory for US federal use and universally adopted.
    • NIST FIPS 203 — ML-KEM (CRYSTALS-Kyber) key encapsulation mechanism; post-quantum standard.
    • NIST FIPS 204 — ML-DSA (CRYSTALS-Dilithium) digital signature; post-quantum standard.
    • RFC 8446 — TLS 1.3 specification, defining the current hybrid encryption standard for internet protocols.
    • ISO/IEC 18033 — International standard series covering encryption algorithms: Part 2 (asymmetric ciphers), Part 3 (stream ciphers), Part 4 (block ciphers).
    • ISO/IEC 27001 / 27002 — Information security management standards requiring encryption controls for sensitive data.
    • ETSI — European Telecommunications Standards Institute issues standards for algorithm agility and lawful interception interfaces.
    • IETF — Internet Engineering Task Force maintains RFCs governing protocol-level encryption: TLS, DTLS, S/MIME, OpenPGP, Noise Protocol.
    • GDPR (EU) / CCPA (California) — Privacy regulations that recognise encryption as a technical safeguard reducing notification obligations after data breaches.
    • PCI-DSS — Payment Card Industry Data Security Standard requiring encryption of cardholder data at rest and in transit.

    Threat Landscape and Limitations

    • Key compromise: encryption provides no protection if the secret key is stolen; Key Management hygiene — secure generation, storage, rotation, and destruction — is paramount.
    • Side-channel attacks: timing, power consumption, electromagnetic emanations, and cache access patterns can leak key material from correct implementations; constant-time algorithms and hardware security modules (HSM) mitigate these.
    • Algorithm deprecation: MD5 and SHA-1 are broken; RC4 and DES are deprecated; 3DES was retired in NIST SP 800-131A Rev. 2. Cryptographic agility — the ability to swap algorithms without redesigning systems — is a design requirement.
    • Harvest now, decrypt later: adversaries capturing today’s encrypted traffic may decrypt it once quantum computers mature; Post-Quantum Cryptography migration addresses this threat.
    • Implementation bugs: memory-safety vulnerabilities in cryptographic libraries (e.g., Heartbleed in OpenSSL) can expose keys or plaintext; formally verified implementations (e.g., HACL*) reduce this risk.
    • Legal and regulatory tension: lawful access demands and “exceptional access” proposals (backdoors) remain controversial; cryptographic consensus holds that algorithmic backdoors are incompatible with strong security.

    Current Landscape (2026)

    • The defining shift is the move to post-quantum encryption: on 13 August 2024 NIST finalised its first three PQC standards — FIPS 203 (ML-KEM, the primary general-encryption key-encapsulation mechanism, formerly CRYSTALS-Kyber), FIPS 204 (ML-DSA signatures) and FIPS 205 (SLH-DSA hash-based signatures) — urging administrators to begin migrating immediately.
    • NIST broadened the portfolio by selecting the code-based HQC (Hamming Quasi-Cyclic) as a backup KEM to ML-KEM on 11 March 2025 (draft standard expected early 2026, final 2027), while FIPS 206 (FN-DSA, based on Falcon) remained in draft as of mid-2026; NIST also published SP 800-227 (Recommendations for KEMs) in September 2025.
    • Real-world deployment accelerated sharply around the hybrid X25519MLKEM768 key exchange: Cloudflare reported on 7 April 2026 that more than half of the human web traffic it processes now uses post-quantum key agreement, up from roughly 2% in early 2024, and Chrome, Edge and Firefox now enable the hybrid handshake by default.
    • Encrypted messaging led the consumer rollout — Apple shipped its PQ3 protocol for iMessage from iOS 17.4 in 2024 (adding ML-KEM to the ongoing ratchet for “Level 3” security), Signal deployed its PQXDH hybrid, and platforms such as Zoom added PQC support; OpenSSL 3.5 now ships a quantum-resistant hybrid KEM.
    • Regulatory and vendor timelines hardened: NSA’s CNSA 2.0 mandates PQC in new national-security acquisitions from 1 January 2027, NIST IR 8547 targets deprecating RSA-2048/ECC P-256 by 2030 and removing quantum-vulnerable algorithms by 2035, and Google announced a 2029 target for completing PQC migration across Chrome, Android and Google Cloud.
    • The frontier is shifting from confidentiality to authentication: Google and others are now prioritising signature migration (ML-DSA/SLH-DSA) to counter “trust now, forge later”, with Let’s Encrypt committing to Merkle Tree Certificates for post-quantum Web PKI (staging late 2026, production 2027).
    • Open challenges as of 2026 include a near-total gap in post-quantum certificates (one measurement study found ~49% of domains support hybrid PQ key exchange but 0% use hybrid PQ certificates), no FIPS 140-3 validated module yet offering PQC in approved mode, and the “harvest now, decrypt later” exposure of long-lived data still under classical RSA/ECC.

    References

    Provenance

    Graph View

    Backlinks