Security implications of novel quantum information technologies

In their early days, new technologies often generate hype cycles that attest to considerable speculation and uncertainty about future application spaces, long-term socio-political impact and security implications. Artificial Intelligence is certainly one of the most striking examples as of late. While many legal and ethical challenges of AI are yet to be met, as Noel Sharkey demonstrates in his recent hbs blog post, developments beyond Machine Learning have also been making considerable progress. Quantum computing and quantum communication in particular mark the beginning of a new computational era which, in many ways, will be considerably superior to “classical” digital computing. 

Quantum information technologies are thus a hot topic at present. Hardly a week goes by when some large multinational company that builds new quantum systems wouldn’t announce a huge breakthrough. Google for instance made headline news in October 2019 when it reported that its Sycamore processor reached quantum supremacy, i.e. the point where a quantum computer can solve a complex problem much faster than the most powerful digital computer ever could. Google’s research scientists describe how their 53-qubit machine had solved a mathematical problem in about 200 seconds; a task they claim the fastest supercomputer manufactured by their competitor IBM, a machine called Summit, would have taken 10,000 years to complete. IBM doubts this but acknowledges the huge progress that Google has made. IBM had introduced Q System One, the world’s first “commercial” quantum computer, in January 2019. 

A radically different computer architecture

What makes a quantum computer so incredibly fast? The chief reason is that it works in entirely new ways. The basic functional principles of a quantum computer are radically different from the binary system of its digital counterpart. Digital computers, such as desktops and laptops, contain billions of tiny transistors to which currents can be applied and that are chained together to build complex circuits. At any point, each transistor is in one of two states: either some current is flowing, which is usually represented by One, or there isn’t, represented by Zero. This principle of “current on/off” is used to create long binary chains of Zeros and Ones that encode more complex information such as words or pixels. Each bit, the basic unit of a digital computer, thus has the binary value 0 or 1. It is important to note that the processing power of a digital computer is a linear function of the number of transistors it contains. 

A quantum computer however is not binary (see figures 1 and 2). Its basic unit is not a bit but a so-called qubit. Photons, electrons or atoms are all possible examples of qubits. Now, a quantum computer utilises quantum phenomena that are not necessarily intuitive. For instance, quite unlike a bit, a qubit can be in more than one state at a time. Following the analogy of the bit, this means that a qubit can be both 0 and 1 at the same time. Moreover, several qubits can be entangled, or chained together in very specific ways so that their computational power won’t grow linearly (as it is the case with bits) but exponentially. For instance, a single qubit can perform two operations simultaneously, two qubits allow for four simultaneous operations, three qubits yield eight operations and so forth: in principle, Google’s 53-qubit Sycamore processor then could perform 2⁵³ operations simultaneously, and this is quite a lot. In theory, a quantum computer of “only” 300 logical qubits could thus perform 2³⁰⁰ operations at the same time – a figure of unimaginable magnitude. Experts estimate that 2³⁰⁰ is roughly the number of all particles in the entire universe. 

It is unlikely that such a machine could ever be built. The rise in qubits on a chip produces a jump in error rates of similar magnitude so that computing quickly becomes wholly unreliable and indeed impossible. Yet the example shows just how much computational power these new systems promise, at least in principle. This is why multinational corporations and governments across the world are engaged in a race to build the first fully functioning quantum computer, which can be expected to enter the stage in about ten years from now. It is important to mention that a quantum computer won’t be universally faster or better than digital computers but only superior concerning some specific (yet important!) applications. Also, the vast costs of the first generations of these machines will make them virtually unusable for private individuals and households. Initial applications that will be relevant for individuals are likely realised through cloud applications, such as superfast data processing or super-secure encryption. 

More importantly, however, quantum computers can run a certain class of algorithms that simply aren’t available to digital computing. And this is a big issue from a security policy perspective. Quantum computers will be able to crack a large class of encryption systems that secure much of our digital communication today. Digital computers find it incredibly hard to factor very great integers that are the product of two (great) prime numbers. While it is not a problem for a digital machine to calculate this product, i.e. multiplying two great prime numbers is easy, it is not easy at all to work “backwards” and find the two prime numbers of which a given integer is the product of. This sluggishness of digital computers is turned into a security advantage and used to generate key pairs for encrypting financial transactions online or to secure videoconferencing. But these great integers won’t be a challenge for quantum computers. A digital computer would need billions, if not trillions of years to crack a 2048-bit encrypted message. In principle, however, a quantum computer would only need a couple of hours to reveal the code. 

Quantum communication: super secure? 

This is not science fiction. It is true to say that the development of these systems will take many years to complete. However, governments, security services and the intelligence community are alarmed. In many instances, confidential data must be stored securely for many decades to come – in ten years from now, quantum computing will become a significant problem as governments will need to work out new ways to best secure their communication channels against attacks from state and non-state actors that have a powerful quantum computer at their disposal. 

This is where quantum communication networks come in. Quantum phenomena such as superposition and entanglement cannot just be used to build superfast computers but also to secure communication, i.e. protect it against eavesdroppers. The principle to do this is called “Quantum Key Distribution” (or QKD). The key idea here is to exploit a basic principle of quantum mechanics: to be taking a measurement of a system invariably means to interrupt the system and change the state of what is to be measured. This means that there is no way for a third party to infiltrate a quantum communication system without setting off an alarm. Exchanging information over QKD provides absolute security and is therefore of great interest to intelligence services across the globe. While it is true to say that QKD itself provides absolute levels of security, it must be noted however that quantum communication is not immune to hacking. Hackers particularly aim to exploit weaknesses of endpoints of such a system, e.g. where it connects to other hardware or digital legacy systems such as laptops. This is a problem as any super secure quantum channel will necessarily be connected to bits of hardware that cannot be quantum protected. For instance, a team of Chinese researchers recently managed to identify the polarisation of photons from reflections of a laser beam outside the box, thus effectively reading the code and cracking the quantum machine. Whilst in principle QKD is secured by the laws of physics, there are quite a few obstacles to overcome to make it work in practice. 

However, considerable progress has been made in quantum communication over the past couple of years. In particular China has established a leadership position in this area (see figure 3). A former analyst with the US security service NSA voiced his concern that for the first time in history, the US could be challenged by a superpower that is technologically superior. Estimates suggest that China invests around $10bn annually in its quantum research centre in Hefei. However, it is difficult to obtain accurate figures about Chinese investment levels. That being said, the considerable work that China is putting into quantum communication becomes obvious in a comparative analysis of patents. Compared to the US, the number of Chinese patents in quantum technologies has more than doubled since 2017. Chinese media report that President Xi Jinping personally considers quantum technologies generally, and quantum communication in particular, of tremendous importance for China to establish global technology leadership. 

In particular China’s launch of its Micius satellite in 2016 attracted significant attention and coverage internationally. Micius shoots photons over vast distances for the purpose of securing information (figure 4). The satellite is quite error-prone, and it only works at night but still, a Chinese-Austrian team of researchers QKD-secured a one-hour video conference between Austria and China, which is a remarkable achievement. China has clearly demonstrated its ambitions to establish a leadership position in quantum communication. Other systems of interest to China include a quantum radar which uses entangled photons to detect even the weakest signal reflections from objects such as US stealth bombers that wouldn’t be detectable by any other form of radar. It seems China wants to target submarines in the Pacific in particular, which would compromise US dominance in this region considerably. 

 

Challenges for security policy 

Combining a superfast quantum computer with a hyper-secure quantum communication network thus presents significant challenges in terms of security policy. In light of China’s progress, Europe and the US have already significantly increased their funding commitment to quantum research. The European Union’s Quantum Flagship funds interdisciplinary research at several European universities to a value of around 1bn euros. In June last year, Germany and six other EU member states reached an agreement to build a basic quantum internet for Europe over the next ten years. This novel hardware network is tasked to protect the Union against cyber espionage and the hacking of sensitive infrastructures such as power grids but shall also secure hospitals and patient data. Somewhat hesitant initially, the US has now changed gear and adopted the National Quantum Initiative Act last year that aims for the US to become a leader in quantum technologies. The Act coordinates federal funding of around $1.5bn. And the cybersecurity division of the British intelligence service GCHQ recommends focusing much more strongly on developing post-quantum cryptography, i.e. encryption systems that are robust enough to hold up to both digital and quantum computers. Hardly making any headline news, a quantum security arms race has been forming as of late that no state actor can afford to lose. 

Against this backdrop, there are several ways for Europe to exercise influence. The European Union could push for international cooperation and agreements that limit the use of quantum technologies to commercial and research activities. To this end, the first step would be to provide a comprehensive risk assessment of plausible military applications of quantum technologies to all stakeholders, not unlike current efforts to inform politicians and regulators how AI can be exploited for military purposes. As nuclear powers such as Russia, the US and China increasingly strike a harsh tone in international affairs, there seems limited scope at present for better multilateral cooperation. Thus focusing on indirect pathways for peaceful cooperation on questions of technology seems a better way forward. Industrial norms and standards, for instance, are tried and tested methods in technology governance that consider commercial as well as security concerns. The degree to which Huawei, the Chinese telecommunications conglomerate, should be involved in building 5G networks in the US and the UK is subject to considerable debate at the moment. These discussions show just how much questions of technical detail are always also questions of international security. For this reason, it can be assumed that China is pushing its quantum communications programme so hard in order to introduce some de facto standards regarding the hardware specifications of novel quantum communication networks. Europe, on the other hand, may try and shape future norms and standards of quantum technologies in cooperation with the International Organization for Standardization (ISO). When it comes to coordinating funding activities over the next couple of years, the European Union should pay close attention to the security dimensions of quantum technologies so that Europe won’t lose considerable ground to China and the US in this new quantum race. 

 

The article was first published on www.boell.de.