How Mathematics Improves The World | Making Digital Trust Possible Between Strangers
Most of the internet is built on a socially ridiculous idea.
You communicate with machines you have never seen.
Those machines may be on the other side of the world.
Your message travels through infrastructure you do not own.
People you do not know may operate parts of that infrastructure.
And still, every day, billions of people expect to:
- log in to a bank;
- buy something online;
- update a phone;
- send private messages;
- sign documents;
- connect to government services;
- download software;
- send business records;
- use cloud systems;
- and know that what arrived is what somebody intended to send.
We do this without personally knowing the engineers who built the servers, the people who maintain the cables, the companies that route the traffic, or the strangers whose devices share the same global network.
Human civilisation needed a new kind of trust.
Not friendship.
Not reputation alone.
Not “I know your face”.
We needed ways for machines to establish useful assurances across open networks.
Mathematics became part of the infrastructure.
Quick Read
Cryptography uses mathematical constructions to protect information and support several distinct kinds of digital assurance. Encryption can help keep information confidential. Message-authentication mechanisms can help detect unauthorised alteration. Digital signatures can help verify that information was signed using a particular private key and that the signed data was not modified afterwards.
Public-key cryptography created an especially important civilisational capability: two parties can establish secure relationships without first sharing the same secret through a private channel. Public information can be distributed widely while a corresponding private key remains secret. That separation supports secure key establishment, digital signatures, certificates and much of modern network security.
The mathematics is powerful, but mathematical security is not the same as complete system security. Keys can be stolen. Software can contain bugs. People can be phished. Random-number generation can fail. Certificate authorities and identity processes can fail. Secure algorithms can be deployed badly.
And cryptography must keep changing. NIST finalised its first three principal post-quantum standards in August 2024—ML-KEM for key establishment and ML-DSA and SLH-DSA for digital signatures—and continues the broader transition toward systems designed to resist attacks from future large-scale quantum computers.
One-sentence answer: Mathematics improves the world by creating verifiable structures for secrecy, integrity, authentication and key exchange, allowing strangers and machines to cooperate across hostile or untrusted networks without requiring complete personal trust first.
The Problem Is Not “Can We Send a Message?”
Sending information is easy compared with sending it safely.
Suppose Alice wants to send Bob a message over a public network.
The first problem is obvious.
Someone else may read it.
So Alice may want confidentiality.
But that is not the only problem.
Someone may alter the message.
So Bob may want integrity.
Someone may pretend to be Alice.
So Bob may want authentication.
A message may be copied and replayed later.
So the system may need freshness or replay protection.
A software vendor may distribute an update to millions of devices.
Those devices need a way to distinguish the genuine update from a malicious imitation.
This is why “security” is not one property.
Different threats require different guarantees.
Confidentiality, Integrity and Authentication Are Different Jobs
These three words are often blurred together.
They should not be.
Confidentiality
Confidentiality asks:
Can unauthorised parties understand the protected information?
Encryption is a major tool for confidentiality.
Integrity
Integrity asks:
Has the information been changed without authorisation?
Cryptographic hashes, message authentication codes and digital signatures can all play roles in detecting alteration, depending on the system.
Authentication
Authentication asks:
Who or what are we interacting with, and what evidence supports that claim?
Digital signatures and challenge-response protocols can contribute to cryptographic authentication.
A secure system may need all three.
And none of them automatically guarantees the others.
The Old Problem: How Do We Share the Secret?
Imagine a perfect secret-key cipher.
Alice and Bob both possess the same secret key.
Alice encrypts with the key.
Bob decrypts with it.
Excellent.
Now ask the awkward question.
How did Alice and Bob securely obtain the same secret in the first place?
If Alice simply emails the secret key to Bob over the same insecure network, an attacker who can read the key can read the later encrypted messages.
Before modern public-key methods, secure key distribution was a major operational challenge. Organisations could use couriers, prearranged secrets, protected physical channels or other trusted mechanisms.
That may work in small or controlled systems.
It does not scale gracefully to a planet of strangers who want secure communication on demand.
The Public-Key Surprise
Public-key cryptography introduced a remarkable architecture.
Instead of one secret key shared by both parties, a user can have a mathematically related key pair:
- a public key that can be distributed;
- a private key that must remain secret.
The details depend on the cryptographic construction, but the civilisational consequence is enormous.
You can publish one part of the system without publishing the power that must remain private.
This breaks an old dependency.
Two parties no longer need to begin every secure relationship by somehow transporting the same secret through a pre-existing private channel.
That does not solve every trust problem.
But it changes the scale at which secure systems can be built.
A Door That Is Easy One Way and Hard the Other
Many public-key systems rely on mathematical problems with a useful asymmetry.
Some operation is computationally practical in one direction but believed to be extremely difficult to reverse without special secret information.
The familiar educational examples involve:
- integer factorisation;
- discrete logarithms;
- elliptic-curve discrete logarithms;
- and, in post-quantum cryptography, different families of hard problems including structured lattice problems.
The security of a real scheme is more complicated than the slogan “one-way function”.
Parameters matter.
Protocol design matters.
Implementation matters.
Attack models matter.
But the broad mathematical insight remains:
Useful security can be built from computational asymmetry.
RSA and the Scale Difference Between Multiplying and Factoring
RSA is historically important and conceptually instructive.
Multiplying two large prime numbers is straightforward for a computer.
Given only their enormous product, recovering the original factors can be far more difficult when parameters are chosen appropriately.
RSA uses number-theoretic structure related to modular arithmetic and this factorisation difficulty.
Students should resist a common oversimplification.
The security claim is not:
Factoring is impossible.
It is a computational claim about practical difficulty at relevant key sizes under known attack methods and computing models.
This is a different kind of certainty from a theorem.
A theorem may establish that a mathematical statement follows from assumptions.
Cryptographic security often also depends on beliefs about the computational difficulty of problems for adversaries.
That distinction becomes crucial when computing technology changes.
Modular Arithmetic: The Clock Mathematics Under the Internet
Students first meet modular thinking informally with clocks.
If it is 10 o’clock and five hours pass, the clock does not say 15.
It wraps around.
10 + 5 is congruent to 3 modulo 12.
That school-friendly idea grows into a deep branch of number theory.
Modern cryptography makes extensive use of arithmetic in finite structures where numbers wrap according to a modulus or related algebraic rules.
This is another example of a recurring theme in Mathematics education.
An idea can begin as a puzzle and end as infrastructure.
Digital Signatures: A Different Job From Encryption
People sometimes describe digital signatures as “encrypting with the private key”.
That slogan is too crude for modern cryptography and can create confusion.
A digital-signature scheme has its own algorithms for key generation, signing and verification.
The conceptual job is this:
- the signer uses a private key to generate a signature associated with data;
- others use the corresponding public key to verify the signature;
- successful verification provides evidence that the signature was generated with the relevant private key and that the signed data has not been modified in a way the scheme accepts.
NIST describes digital signatures as providing assurance about the claimed signatory and whether information was modified after signature generation.
This is extraordinarily useful for software distribution, documents, protocols and many forms of machine-to-machine trust.
A Signature Does Not Tell You Whether the Signer Is Good
This boundary matters.
Suppose a digital signature verifies correctly.
What does that prove?
It provides cryptographic evidence tied to a key and the signed data.
It does not automatically prove that:
- the human behind the key is honest;
- the key was never stolen;
- the signer understood the document;
- the signed statement is factually true;
- the software is harmless;
- the institution issuing an identity certificate performed perfect verification.
Cryptographic trust is scoped.
This is a powerful general lesson.
Whenever a system says “verified”, ask:
Verified what, exactly?
Good Mathematics helps specify the guarantee instead of letting a reassuring word become bigger than the evidence.
Hash Functions: Fingerprints for Data, With Important Caveats
A cryptographic hash function takes an input and produces a fixed-length output often called a hash or digest.
Good cryptographic hash functions are designed to have properties that make them useful for integrity and many protocol constructions.
A tiny change in the input should typically produce a very different digest.
It should be computationally difficult to recover an arbitrary original message from its digest.
It should be computationally difficult to find two distinct inputs with the same digest under the intended security assumptions.
This makes hashes useful as compact representatives of data.
Digital-signature systems commonly sign a representation derived from the message rather than processing an arbitrarily large file naively.
Again, metaphor helps but must be bounded.
A hash is often compared with a fingerprint.
Useful.
But a fingerprint is biological evidence; a cryptographic hash is a deterministic mathematical function output.
The analogy transfers the idea of a compact identifier.
It should not erase the difference.
Certificates: The Mathematics Needs an Identity Story
Suppose a website sends you a public key.
Excellent.
Whose public key is it?
If an attacker intercepts your connection and substitutes their own key, beautiful mathematics can protect your conversation with the wrong party.
This creates a binding problem:
How do we connect a public key to an identity or domain in a way other systems can evaluate?
Public-key infrastructures use certificates and trusted issuers to help bind keys to identities or names under defined procedures.
This adds an institutional layer around the mathematics.
The cryptography can say:
this certificate was signed by a key trusted for this purpose.
The surrounding system must decide:
- which authorities are trusted;
- how identities are validated;
- how certificates expire;
- how compromised credentials are revoked;
- how software handles errors;
- what happens when trust anchors themselves fail.
Digital trust therefore has layers.
Mathematics is a foundational one.
It is not the only one.
The Browser Padlock Was Never a Moral Certificate
People used to be told to look for a padlock icon in the browser.
The advice was useful only within limits.
An encrypted connection can protect communication with a site.
It does not prove that the site’s business model is ethical.
It does not prove that every claim on the site is true.
It does not prove that the operator will treat your data well after receiving it.
Transport security is not character judgement.
This boundary is important because humans naturally inflate trust signals.
A green icon becomes “safe”.
A verified signature becomes “true”.
A certificate becomes “good company”.
Mathematical literacy should make us more precise than that.
Symmetric Cryptography Still Does Most of the Heavy Lifting
Public-key cryptography is glamorous because the idea is surprising.
But modern secure communication usually combines cryptographic tools.
Public-key techniques can help establish shared secrets or authenticate parties.
Fast symmetric algorithms then protect large amounts of data efficiently.
This hybrid architecture is another example of good engineering.
Do not ask one tool to do everything.
Use each tool where its properties are strongest.
Randomness: Security Can Fail Before the Equation Begins
Cryptographic systems often depend on random or unpredictable values.
Keys need sufficient unpredictability.
Nonces may need uniqueness or unpredictability depending on the construction.
Protocol challenges may rely on fresh random values.
If the random-number generator is weak, a mathematically strong algorithm can collapse operationally.
This is a recurring principle of real-world Mathematics:
The strength of a theorem does not rescue a system whose inputs violate the assumptions.
Security is a chain.
The equation is one link.
Side Channels: The Secret Can Leak Around the Mathematics
Suppose a cryptographic algorithm is mathematically sound.
An attacker may still learn secrets from the way the algorithm is implemented.
Timing differences can leak information.
Power consumption can leak information.
Electromagnetic emissions can leak information.
Error messages can leak information.
Memory handling can leak information.
These are called side-channel or implementation attacks in various forms.
The lesson is sobering.
You can lose a mathematical secret through physics.
Once again, disciplines meet.
Phishing: Mathematics Cannot Stop You Handing Over the Key
Cryptography can make a key computationally difficult to derive.
It cannot stop a human being from revealing a password to a convincing impostor.
It cannot stop an employee from approving a malicious request if the surrounding process is poor.
It cannot stop malware from using a legitimate key after compromising the device that stores it.
This is why cybersecurity is not reducible to cryptography.
Mathematics can make certain attacks extraordinarily difficult.
Attackers then look for cheaper routes.
The strongest door is irrelevant if someone is persuaded to open it.
Why Quantum Computing Changes the Security Conversation
Modern public-key cryptography relies on computational problems believed to be hard for ordinary classical computers at suitable parameters.
Large-scale fault-tolerant quantum computers, if built with sufficient capability, would change the difficulty of important mathematical problems underlying widely deployed public-key systems.
That creates a long-horizon infrastructure problem.
We do not wait until a cryptographic foundation fails before beginning to replace global infrastructure built on it.
NIST began a multi-year process to evaluate and standardise post-quantum public-key algorithms.
On 13 August 2024, NIST released three principal standards:
- FIPS 203 — ML-KEM, a module-lattice-based key-encapsulation mechanism for establishing shared secret keys;
- FIPS 204 — ML-DSA, a module-lattice-based digital-signature standard;
- FIPS 205 — SLH-DSA, a stateless hash-based digital-signature standard.
NIST subsequently selected HQC for future standardisation as an additional key-establishment option, and its post-quantum programme continues to evaluate additional digital-signature candidates.
This is mathematics responding to a future threat before the future fully arrives.
What Is a Lattice Doing in Cryptography?
The word lattice may make you imagine a garden fence.
In mathematics, a lattice can refer to a regular discrete structure generated by integer combinations of basis vectors in a vector space.
In high dimensions, lattice problems can become extremely rich and difficult.
Post-quantum schemes such as ML-KEM and ML-DSA use structured algebraic versions of lattice-related problems rather than simply drawing a high-dimensional grid and hiding a point in it.
The exact security constructions are advanced.
But the educational point is accessible.
When one family of mathematical hard problems may become vulnerable to a new computing model, civilisation can search for different hard problems whose structure is believed to remain resistant.
Mathematical diversity becomes security resilience.
Post-Quantum Does Not Mean Quantum Encryption
This terminology often confuses readers.
Post-quantum cryptography generally refers to cryptographic algorithms designed to run on ordinary computing platforms while resisting attacks from adversaries who may possess powerful quantum computers.
It is different from quantum key distribution and other approaches that use quantum physical systems directly.
The distinction matters because post-quantum algorithms can be integrated into conventional digital infrastructure without requiring every endpoint to become a quantum device.
The Hardest Part May Be Migration
Inventing a secure algorithm is not the same as replacing a civilisation’s cryptographic infrastructure.
Old algorithms are embedded in:
- operating systems;
- network protocols;
- hardware security modules;
- industrial systems;
- embedded devices;
- financial infrastructure;
- government systems;
- identity platforms;
- software-update systems;
- long-lived archival workflows.
Some devices are difficult to update.
Some systems are expected to remain deployed for decades.
Some encrypted information must remain confidential for many years.
This creates the “harvest now, decrypt later” concern: an adversary can store encrypted communications today in the hope of decrypting them in the future if the relevant cryptography becomes breakable.
So the transition has to begin before the threat is fully operational.
This is a striking example of mathematics changing infrastructure planning decades ahead.
Cryptographic Agility: Designing for the Day Today’s Answer Is No Longer Enough
A system that hard-codes one algorithm everywhere becomes painful to update.
Cryptographic agility means designing systems so algorithms, parameters and keys can be changed when security requirements evolve.
This is not merely a cybersecurity detail.
It is a philosophy of infrastructure.
Build for correction.
Assume that future knowledge will change some current choices.
Do not confuse “best available today” with “correct forever”.
That is mathematical humility turned into engineering architecture.
The Mathematics of Trust Is Really the Mathematics of Limited Guarantees
Human trust is broad.
We say:
“I trust her.”
Cryptographic trust is narrower and, in some ways, healthier.
We ask:
- Do I have evidence that this message was not altered?
- Do I have evidence that the signer possessed a particular private key?
- Do I have a protected shared key with this endpoint?
- Which authority binds this key to this domain?
- Which algorithms and parameters support the guarantee?
- What assumptions would make the guarantee fail?
This is not emotional trust.
It is scoped assurance.
Mathematics improves the world by making some forms of trust smaller, clearer and testable.
A Simple Thought Experiment: The Locked Box With an Open Padlock
Here is a physical metaphor for public-key encryption.
Bob sends the world an open padlock.
Anyone can place a message in a box and lock it using Bob’s padlock.
But only Bob has the private key that opens the lock.
The metaphor captures one useful asymmetry:
public capability to protect something for Bob, private capability to recover it.
But the metaphor has limits.
Real public-key encryption uses mathematical algorithms, not physical padlocks.
Protocols may use public-key methods to establish symmetric session keys rather than encrypt every message directly.
And a padlock sent through the post still raises the identity question:
How do you know the padlock really belongs to Bob?
That last question leads straight back to certificates and authentication.
A Second Thought Experiment: Signing a Million Copies Without Touching Them
Imagine a software company releasing an update to ten million devices.
The company cannot physically visit each device.
Every device needs a way to ask:
Was this update signed by a key I am configured to trust for updates, and has the signed content changed?
Digital signatures allow the same mathematical verification process to scale across enormous populations of devices.
This is where abstract algebra becomes logistical power.
One signing operation can create evidence that millions of independent machines can verify.
Primary Mathematics: The First Seeds of Cryptographic Thinking
A Primary student does not need modern cryptography.
But some useful habits begin early.
Patterns.
Remainders.
Factors and multiples.
Prime numbers.
Logical conditions.
Checking whether a rule works for every case.
Separating a secret from a public instruction.
A child who plays with a Caesar cipher should not be told they have learned secure cryptography.
They have learned something better suited to their level:
information can be transformed according to rules, and the security of a rule depends on what an adversary can infer.
Secondary Mathematics: Number Structure Becomes More Visible
Secondary students gain stronger algebra and number reasoning.
They learn powers and indices.
Functions.
Algebraic manipulation.
Proof-like reasoning.
Probability.
Combinatorics in various forms.
These are not “cryptography chapters”, but they enlarge the student’s ability to reason about structured transformations and huge search spaces.
Advanced Mathematics: Where the Security Machinery Lives
Higher study can lead into:
- number theory;
- abstract algebra;
- finite fields;
- elliptic curves;
- probability;
- information theory;
- complexity theory;
- coding theory;
- lattice theory;
- discrete mathematics;
- algorithm analysis.
At that level, cryptography becomes a conversation among pure mathematics, algorithms, probability, systems engineering and adversarial thinking.
This is another reason the pure/applied divide can mislead students.
Some of the most abstract mathematics can later become practical infrastructure.
The Adversary Changes the Mathematics
A normal algorithm is often judged by whether it produces the correct output efficiently.
A cryptographic algorithm lives in a harsher world.
Someone is deliberately trying to break it.
The attacker can choose inputs.
Observe outputs.
Exploit implementation details.
Combine information across many sessions.
Search for the weakest assumption.
This means cryptographic Mathematics is shaped by adversarial questions:
- What information does this leak?
- What if the attacker chooses the message?
- What if the same nonce is reused?
- What if the key is partly known?
- What if the attacker owns a quantum computer?
- What if the implementation takes slightly different time on different secret values?
The mathematics improves because the opponent is allowed to be clever.
Security Proofs Are Not Magic Shields
Modern cryptography often uses formal security models and reductions.
Very roughly, a proof may show that if an attacker could break a cryptographic construction under a defined model, that attacker could also solve some underlying mathematical problem believed to be hard.
This is powerful.
But every proof has assumptions and a model.
A proof about an algorithm does not prove that a particular device has no side-channel leak.
A proof about a protocol does not prove that users will never reveal their credentials.
A proof relative to a computational hardness assumption does not guarantee that no future algorithm will challenge that assumption.
Once again:
Strong Mathematics gives strong conclusions inside clearly stated boundaries.
How Mathematics Improves the World Here
1. It lets secrecy scale
Modern encryption allows sensitive information to travel through networks that cannot be treated as physically private.
2. It lets integrity scale
Cryptographic verification lets systems detect whether protected data has been altered.
3. It lets signatures scale
A signer can create evidence that many independent recipients can verify without sharing the private signing key.
4. It separates public coordination from private power
Public keys can be widely distributed while private keys remain controlled.
5. It makes assumptions inspectable
Cryptographic design forces engineers to state threat models, key sizes, algorithms and security properties rather than relying only on vague reassurance.
6. It allows civilisation to prepare for future attackers
Post-quantum standardisation shows mathematical infrastructure being replaced before old foundations are expected to fail at scale.
What Mathematics Does Not Do
Mathematics does not make every encrypted service ethical.
It does not prevent users from being deceived.
It does not protect a private key that malware can steal.
It does not guarantee a certificate authority’s identity checks were perfect.
It does not remove bugs from software.
It does not make data safe after an authorised recipient mishandles it.
It does not eliminate the need for law, governance, engineering, secure hardware, operational discipline and user education.
Digital trust is a system.
Cryptography is one of its strongest structural members.
For Parents: Why This Is Mathematics, Not Just Computing
A child may look at a phone and think technology is mostly coding.
Coding matters.
But beneath secure digital systems is a long mathematical inheritance.
Prime numbers once looked like objects of pure curiosity.
Modular arithmetic looks like playful clock arithmetic at first.
Abstract algebra can look impossibly detached from daily life.
Then civilisation builds banking, software distribution, secure communications and digital identity systems on mathematical structures from those worlds.
That is why education should not define usefulness too narrowly or too early.
We do not always know which abstract idea will become tomorrow’s infrastructure.
Frequently Asked Questions
What is cryptography?
Cryptography is the study and construction of methods for protecting information and communication against defined adversarial threats. Depending on the scheme, goals can include confidentiality, integrity, authentication, digital signatures and secure key establishment.
What is public-key cryptography?
Public-key cryptography uses related public and private keys. The public key can be distributed, while the private key remains secret. Different schemes use this architecture for key establishment, encryption, signatures or authentication.
Is a digital signature the same as a handwritten signature?
No. They serve analogous evidentiary purposes in some contexts, but a digital signature is a cryptographic construction generated and verified using algorithms and keys. Its precise legal meaning depends on jurisdiction and process.
Does a valid digital signature prove a message is true?
No. It can provide evidence about the signature, key and integrity of signed data. A person can cryptographically sign a false statement. Cryptographic authenticity and factual truth are different properties.
Why do we need post-quantum cryptography?
Powerful quantum computers would threaten important public-key systems based on factorisation and discrete-logarithm problems. Post-quantum cryptography uses different mathematical foundations designed to remain secure against both classical and quantum adversaries under current knowledge.
What are NIST’s main post-quantum standards?
NIST finalised FIPS 203 for ML-KEM, FIPS 204 for ML-DSA and FIPS 205 for SLH-DSA on 13 August 2024. NIST continues additional standardisation work, including HQC as an additional key-establishment algorithm and further digital-signature candidates.
Can strong encryption make me completely secure online?
No. Security also depends on endpoint protection, identity processes, software quality, key management, user behaviour, protocol design, updates and governance. Attackers often target the cheapest weak link rather than attacking the strongest mathematics directly.
Why should a school student care about number theory?
Because number theory is a powerful example of abstract Mathematics later becoming technological infrastructure. It also trains proof, structure and pattern reasoning even when the student never becomes a cryptographer.
Continue Through eduKateSG
Continue with How Mathematics Works, then explore the other articles in this series: Seeing Inside Without Cutting Open, Sending a Spacecraft to Somewhere That Is Moving, and When the Right Match Can Save a Life.
Sources and Further Reading
- NIST, Digital Signatures. NIST’s overview of the assurance digital signatures provide and current standards.
- NIST, Approval of FIPS 203, 204 and 205. The August 2024 announcement for the first three principal post-quantum standards.
- NIST, Post-Quantum Cryptography. Current programme information, standards and migration guidance.
- NIST, FIPS 204 — Module-Lattice-Based Digital Signature Standard. The formal ML-DSA standard and related documentation.
Final Thought: Trust Does Not Have to Begin With Knowing the Person
For most of human history, trust was local.
You knew the merchant.
You recognised the seal.
You knew the messenger.
You belonged to the village.
You could ask someone who knew someone.
Digital civilisation shattered that scale.
Now a Singapore student can connect to a server in another continent in less time than it takes to say the server’s name.
Personal familiarity cannot scale that quickly.
So mathematics supplies something different.
Not complete trust.
Specific assurances.
This key matches that signature.
This data has not changed.
This shared secret can be established across a public channel.
This algorithm is designed around a defined hard problem.
This guarantee holds under these assumptions.
That precision is one of Mathematics’ gifts to the world.
It lets strangers cooperate without pretending they are not strangers.