Steal Now, Decrypt Later: Why Quantum Computing Means Encryption Has to Change Now

Norman Willox and Tom Patterson are global cybersecurity experts.

Nobody can say with certainty where quantum computing will stand a few years from now, according to cybersecurity experts Norman Willox and Tom Patterson. What they do say is that governments, universities and companies across the world are already making large gains in the race for "quantum advantage", and that the rest of us should be asking what those players, and our adversaries, are doing about it.

Once quantum advantage arrives, the authors expect attackers to reach it without building a super-cooled laboratory of their own. The more likely route, in their view, is a cloud service, in the same way that sophisticated ransomware tooling is now within reach of any criminal who owns a computer and a credit card.

On the defensive side, they point out that important pieces of quantum resistance and quantum encryption already exist, quantum communication has started, and quantum clouds are starting to appear for sensitive work. Their conclusion is that governments and businesses should begin preparing immediately, because adversaries have already begun.

Data is being stolen today to be read tomorrow

Willox and Patterson describe the danger to governments, critical infrastructure and organisations of every size as real, a question of mathematics rather than speculation. They say it has already started, in the form of a new pattern of hostile behaviour known as "steal now, decrypt later" (SNDL). Attackers taking this approach copy large quantities of sensitive encrypted data that they have no way of reading yet, betting that the quantum machines they expect to have soon will open it.

The authors add that quantum-assisted code breaking will be usable years before quantum systems as a whole reach maturity. For them, that is one more reason why the most valuable information needs protecting at once.

They cite the weight Washington has placed on the issue. On 4 May 2022 the President of the United States signed a National Security Memorandum and an Executive Order designed to secure the country's competitive edge in quantum information science (QIS) and to reduce the risks quantum computers pose to national cyber, economic and security interests. It was, the authors note, the fourth such measure that year alone.

The road to "Y2Q"

Today's public key encryption, they explain, rests on an old assumption: that even the fastest computers would need many millions of years to factor large prime numbers. Each time machines became a little quicker, the answer was simply to lengthen the key and keep the assumption intact.

Over the past decade, however, quantum computing has moved from science fiction to science fact, and the authors warn that the world is drawing steadily nearer to "Y2Q". That is the point at which a quantum computer can run Shor's algorithm and read everything that has ever been encrypted, whatever the key length. Stronger encryption has to be invented before that day, and it also has to be adopted, distributed, installed and maintained around the globe ahead of it. Because that process takes years, they argue it needs to begin now.

Both authors back the Quantum Computing Cyber Preparedness Act, a bipartisan bill put before the US House of Representatives in April 2022 that aims to accelerate, strengthen and regulate quantum cybersecurity. They welcome it for drawing attention to how badly adversarial quantum decryption could damage government files and communications. Their criticism is that it leaves out the 16 critical infrastructure sectors run by the private sector, where the same need exists, and they believe Congress could do more to shield private companies and businesses from a threat that is emerging and may be close.

Why the private sector cannot wait

Roughly 85% of US critical infrastructure is in private hands, the authors say. They ask readers to picture health records exposed, bank balances wiped, transport halted and power cut, and point out that every one of these sectors depends on trusted encryption for even its most basic functions. The federal government is also supported by a very large defence and security industrial base holding extensive sensitive government and industry information, and in their view protecting those supply chains matters as much as protecting the agencies.

To guard against criminals, terrorists or intelligence services putting quantum computing to use, Willox and Patterson believe every part of government and every critical infrastructure sector should start a four-step process straight away:

1. Take a full inventory of where the organisation relies on encryption, recording the algorithm, key distribution, provider and partners involved in each case.
2. Start making encryption "agile", so that it can be changed more easily later.
3. Adopt the newest encryption on offer today, such as Messaging Layer Security (MLS), which was built to withstand aggressive collection of communications and collaboration traffic, along with quantum-generated shared keys for symmetric algorithms.
4. Study and test both the "quantum resistance" algorithms that are NIST candidates (obtainable through the providers found in the inventory) and the newer "quantum encryption" systems, which apply quantum physics that is available today to random numbers, keys and more, giving provably secure encryption now with some existing algorithms.

The authors see these steps as what success depends on, both now and later. A strategy for quantum proofing is required today, they say, and finding the right people, specialists, partners, products and tools to carry it out and keep it going will be critical.

Not only a government problem

It is easy to understand why many assume adversarial quantum computing is something only governments need to worry about, the authors write, but they argue it could touch every person and every business. Everyone holds secrets, intellectual property or sensitive information on which their business or life depends, and everyone is exposed once it leaks. Ransomware has already shown how quickly the most advanced cyber weapons end up with criminals. They invite readers to consider which confidential data they trust systems to protect, and whether they would choose a product able to keep it safe into the future.

Commercial data, they note, far outweighs government data in volume. With digital data and AI growing explosively, the value of information is rising exponentially, and the threat and the risk are rising with it.

A large share of that data is temporal, tied to a particular moment, so a future adversary decrypting it would make little difference. The HTTPS connection a reader is probably using to view an article is one example. Other data is nontemporal, meaning its value does not fade with time, and has to stay protected for a long time to come. The authors point to news reports of a foreign power copying a database of sensitive background information on people holding US Government security clearances, their own included. That, they say, is nontemporal data they would have liked to stay secure for at least the rest of their lives.

What quantum computing will make possible

Having raised the alarm, Willox and Patterson are just as keen to set out the upside. Quantum computing, they write, offers computation that is both faster and fundamentally different. Entirely new problems can be formulated and solved, communications can become instantaneous, universal and secure, and remote sensing will become real, among much else. Outside code breaking, fintech, pharmaceuticals, logistics, communications, space, climate and data analytics are all working actively to put the coming machines to use.

The authors recall Albert Einstein's famous description of the quantum world as "spooky". Knowing what is known now, they find quantum computing exhilarating instead, and expect it to carry people to intellectual places no one has yet imagined and to solve problems once thought unsolvable. Their closing line: "Welcome to the future."