Public-Key Cryptography: From Classified Prehistory to Diffie-Hellman and RSA
A careful chronology of public-key ideas at GCHQ, Diffie and Hellman's 1976 publication, Merkle's contribution, RSA, signatures, and attribution.
In 1976, Whitfield Diffie and Martin Hellman published “New Directions in Cryptography,” introducing public-key cryptography and a practical key-agreement construction to the open research community. Ralph Merkle’s puzzles and related work were important inputs. In 1977, Ronald Rivest, Adi Shamir, and Leonard Adleman developed the RSA system, published in 1978. Decades later, declassification revealed earlier classified concepts at Britain’s GCHQ.
The key-distribution problem drove the open work
Traditional symmetric encryption requires communicating parties to share a secret key. In a large network, secure key distribution and storage become an organizational problem before any message is encrypted. Diffie and Hellman proposed separating encryption and decryption capabilities so a public value could be distributed openly while a private value remained secret.
Their paper described a public-key encryption concept and gave an exponential key-exchange method based on modular arithmetic. Two parties publish values derived from private exponents and compute the same shared secret without transmitting it directly. The construction now commonly called Diffie-Hellman does not authenticate the parties by itself; an active intermediary can establish separate keys unless signatures, certificates, pre-shared authentication, or another mechanism binds identities.
Merkle independently investigated secure communication over insecure channels and developed puzzles with a work-factor gap. Diffie and Hellman credited Merkle’s influence; modern historical naming sometimes uses Diffie-Hellman-Merkle key exchange to make that role visible. Attribution conventions vary, so primary papers and explicit contribution descriptions are preferable to one winner’s label.
RSA supplied encryption and signatures from one trapdoor idea
Rivest, Shamir, and Adleman used exponentiation modulo a composite integer whose factorization is secret. Their 1978 paper described a method for obtaining digital signatures and public-key cryptosystems. Security depends on parameter choices, hard mathematical problems, and safe encoding—not merely applying the textbook exponent operation.
Textbook RSA is deterministic and algebraically malleable. Modern encryption and signature use standardized padding/encoding such as OAEP and PSS, vetted libraries, appropriate key sizes, and side-channel-resistant implementations. The history of the core idea should not be turned into current cryptographic implementation advice.
Public-key signatures addressed another Diffie-Hellman goal: a private operation can produce evidence verified with public information. Real identity still needs a trust system—certificates, key directories, transparency logs, personal verification, or an organizational authority. Mathematics does not decide who controls a named key.
Classified work complicates “first”
GCHQ later disclosed that James Ellis conceived non-secret encryption in 1970, Clifford Cocks found an RSA-like construction in 1973, and Malcolm Williamson discovered a key-agreement method in 1974. The work remained classified and did not influence the open community’s 1970s publications according to the disclosed history.
This produces two legitimate chronologies: earliest known internal conception and first open publication that enabled independent scrutiny and adoption. Cocks preceded RSA internally; RSA independently produced, published, analyzed, and became deployable. Diffie, Hellman, and Merkle created the open conceptual framework and key-exchange line while unaware of GCHQ’s work.
Publication changed the research system
Open papers let mathematicians and engineers challenge assumptions, develop attacks, standardize encodings, and build protocols. Patents and export controls shaped early commercialization in the United States, while academic work expanded elliptic-curve and other public-key families.
The durable breakthrough was not “encryption without secrets.” Private keys remain secrets, implementations leak, algorithms age, and authentication remains necessary. It was a new division of cryptographic capability: public distribution for establishing or verifying, private control for decrypting or signing. Precise history makes the same point as precise security engineering—separate what each contribution and each key actually does.
Related:
- The Transistor at Bell Labs: From Point Contact to a Reproducible Junction Device
- January 1, 1983: What ARPANET’s TCP/IP Flag Day Actually Changed
Sources: