How Encryption Keeps Data Safe

01 The Ancient Art of Hidden Writing

Encryption is older than the alphabet. The earliest known example of deliberate symbol substitution — the act of transforming readable text into something only an initiated reader can understand — appears in the tomb of Khnumhotep II, an Egyptian nobleman who lived around 1900 BC. The substitution was not a true cipher in the mathematical sense, but it established the principle: information could be made unintelligible to those who did not possess the key.

The first systematic military cipher was the Caesar cipher, used by Roman forces to protect messages carried by couriers through hostile territory. The mechanism is simple: shift each letter of the alphabet by a fixed number of positions. With a shift of three, A becomes D, B becomes E, and the word "ATTACK" becomes "DWWDFN." Anyone who knows the shift can decrypt; anyone who does not sees only nonsense. For a world without computers, this was adequate — the number of possible shifts was small, but the ability to test them all was limited by manual labour.

The Caesar cipher fell to frequency analysis, a technique developed around 800 AD by the Arab polymath al-Kindi. By counting how often each letter appears in the ciphertext and comparing to the known frequency of letters in the language, a cryptanalyst can deduce the shift without trying every possibility. This launched a centuries-long arms race between cipher makers and code breakers — a contest that would eventually transform cryptography from an art practiced by monks and diplomats into a mathematical discipline at the heart of computer science.

02 Symmetric Encryption: One Key, Two Parties

In symmetric encryption, the sender and receiver share a single secret key. The key is used both to scramble the plaintext into ciphertext and to unscramble it back again. Every cipher from the Caesar shift to the Enigma machine to the Advanced Encryption Standard operates on this principle: the same key opens and locks the box.

The AES algorithm, approved by the US National Institute of Standards and Technology in 2001, is the workhorse of modern symmetric encryption. It processes data in 128-bit blocks and supports key sizes of 128, 192, or 256 bits. The 256-bit variant — AES-256 — is approved by the NSA for protecting top-secret classified information. Its security rests not on the algorithm being secret but on the key being impractically large to guess. A 256-bit key has 2^256 possible values, a number larger than the estimated number of atoms in the observable universe. Even if every computer on Earth had been trying one billion keys per second since the Big Bang, they would have examined an infinitesimal fraction of the keyspace.

The weakness of symmetric encryption is not the cipher itself but the problem of key distribution: how do two parties who have never met agree on a shared secret without anyone overhearing? For millennia, this required a trusted courier or a pre-arranged codebook. The solution that transformed cryptography — and with it, the entire architecture of digital communication — arrived in 1976.

03 Public-Key Cryptography: The Key Revolution

In 1976, Whitfield Diffie and Martin Hellman published a paper that solved the key-distribution problem with a mathematical trick. Instead of one shared key, each party has two: a public key that anyone can know, and a private key that is never shared. A message encrypted with the public key can only be decrypted with the corresponding private key. The public key can be distributed openly — printed on a business card, posted on a website — because possessing it does not help an attacker decrypt messages.

The mechanism rests on one-way functions: mathematical operations that are easy to perform but computationally infeasible to reverse. The RSA algorithm, published in 1977 by Ron Rivest, Adi Shamir, and Leonard Adleman, uses the difficulty of factoring large composite numbers. Multiplying two large prime numbers together is trivial; given only the product, finding the original primes takes exponentially longer. With a 2,048-bit RSA key, the product has over 600 digits. Factoring it with the best known classical algorithm would require more computing time than the expected lifespan of the universe.

Public-key cryptography also enables digital signatures. A sender can encrypt a hash of their message with their private key; anyone with the sender's public key can verify that the signature was produced by the private key holder and that the message has not been altered. This is the foundation of certificate authorities, code signing, and the chain of trust that makes web browsing secure. When your browser displays a padlock icon, it has verified a digital signature chain back to a trusted root certificate — a process that happens in milliseconds but depends on decades of cryptographic research.

04 The Hybrid System: How HTTPS Actually Works

Every time you visit a secure website, your browser and the server perform a cryptographic handshake that combines both symmetric and public-key cryptography. The reason is performance: public-key operations are computationally expensive — roughly 1,000 times slower than symmetric operations. You would not want to encrypt an entire web page with RSA; you would still be waiting for the first image to load.

The TLS protocol, which underpins HTTPS, solves this with a hybrid approach. The client and server use public-key cryptography to authenticate each other and to exchange a shared symmetric key. Once the symmetric key is established — a process taking milliseconds — all subsequent data is encrypted with AES or ChaCha20, which can encrypt bulk data at gigabits per second. The public-key layer provides authentication and key exchange; the symmetric layer provides fast, efficient confidentiality.

This hybrid architecture is why encryption is invisible to most users. The handshake happens in the background; the browser displays a padlock and moves on. But the underlying machinery involves certificate validation, key exchange protocols like Diffie-Hellman, cipher negotiation, and forward secrecy — the property that even if a server's long-term private key is compromised in the future, past sessions encrypted under ephemeral keys remain secure. Every secure connection you make today benefits from four decades of accumulated cryptographic engineering, operating in milliseconds.

05 Hashing: The One-Way Function

Not all cryptography is about secrecy. Some of it is about integrity and authentication — proving that data has not been altered and that it came from whom it claims. The tool for this is the cryptographic hash function: a mathematical operation that takes any input and produces a fixed-size output called a hash, with three critical properties. First, it is deterministic: the same input always produces the same hash. Second, it is one-way: given a hash, it is infeasible to find the input that produced it. Third, it is collision-resistant: it is infeasible to find two different inputs that produce the same hash.

The SHA-256 algorithm, part of the SHA-2 family published by the NSA in 2001, produces a 256-bit hash from any input — a single letter or an entire terabyte of video. The hash is a fingerprint: change one bit of the input and the entire output changes unpredictably. This makes hashes ideal for verifying data integrity, storing passwords (in hashed form, so the server never knows the plaintext), and constructing blockchains, where each block's hash incorporates the previous block's hash, creating a tamper-evident chain.

Password hashing deserves special mention. When you create an account on a well-designed service, your password is not stored. Instead, a salt — a random value unique to your account — is appended to your password, and the combination is hashed. Only the salt and the hash are stored. When you log in, the service recomputes the hash with the stored salt and checks if it matches. Even if the database is stolen, the attacker obtains only hashes — which must be cracked one by one, with each attempt requiring a full hash computation. Slow hashing functions like bcrypt and Argon2 deliberately make each computation expensive, so that a billion-password attack that would take seconds against SHA-256 takes years against Argon2.

06 The Quantum Threat

Every cryptographic system discussed so far relies on a computational assumption: that certain mathematical problems are hard for any computer. Factoring large numbers, computing discrete logarithms, finding hash collisions — all are believed to be intractable on classical computers. But this assumption does not hold for quantum computers. In 1994, Peter Shor published an algorithm that can factor large numbers in polynomial time on a quantum computer, collapsing the security of RSA and Diffie-Hellman. A sufficiently large quantum computer would break the public-key cryptography that protects virtually all digital communication.

As of 2026, no quantum computer large enough to break RSA-2048 exists. Current quantum machines have hundreds of physical qubits; breaking RSA-2048 would require thousands of logical qubits, each built from many physical qubits with error correction. Estimates of when this becomes feasible range from ten to thirty years, but the uncertainty is real. The threat is not just future decryption — it is the harvest-now-decrypt-later attack, where adversaries record encrypted traffic today and store it for the day a quantum computer becomes available.

The response is post-quantum cryptography (PQC): new algorithms whose security does not depend on problems that quantum computers can solve efficiently. NIST began a standardization process in 2016 and published the first PQC standards in 2024: ML-KEM (based on lattice problems) for key encapsulation and ML-DSA for digital signatures. These algorithms are being integrated into TLS and other protocols, beginning the largest cryptographic migration in history. The transition will take years — every system, library, and certificate must be updated — but it is underway, and the alternative is a future in which every secret ever encrypted under the old assumptions becomes readable.

07 The Invisible Infrastructure

Encryption is the most consequential technology that most people never think about. Every bank transfer, every text message, every streaming session, every software update, every election system, every medical record — all depend on cryptographic protocols that operate silently in the background. The padlock in the browser address bar is the visible tip of an iceberg comprising number theory, computational complexity, hardware security modules, certificate authorities, and decades of peer-reviewed cryptanalysis.

The fragility of this infrastructure is real. A single flaw in a cryptographic implementation — like the Heartbleed bug in OpenSSL, discovered in 2014 — can expose millions of servers to attack. A weakness in a random number generator can undermine an entire protocol, as the Dual_EC_DRBG scandal demonstrated. A mathematical breakthrough — a new factoring algorithm, a faster quantum computer — could render entire classes of encryption obsolete overnight. Security is not a state but a process: a continuous effort to stay ahead of both mathematical discovery and adversary capability.

And yet, for all its fragility, the system works. The vast majority of digital communication is encrypted, and the vast majority of that encryption is never broken. This is not luck; it is the product of open standards, public review, and a research community that has spent fifty years stress-testing every algorithm before it is trusted with real data. Encryption is proof that a mathematical idea — that some problems are too hard to solve — can become the foundation of a global infrastructure. The next time you see a padlock, consider the invisible architecture it represents: the accumulated genius of Shannon, Diffie, Hellman, Rivest, Shamir, Adleman, and thousands of others who turned the ancient art of hidden writing into the engineering of digital trust.