Why quantum technologies need AI to succeed

Experts at the OECD’s Global Forum on Technology identified AI’s support for quantum development as one of the most tangible synergies between the two fields. This article examines AI’s role across the three main branches of quantum technologies: quantum computing, quantum sensing, and quantum communication.

The first blog in this series introduced quantum technologies and explored their complementarity with AI. This instalment looks at how AI is contributes to the development of quantum systems and applications.

Quantum computers: Not just one of a kind

There are several types of quantum computers. While large-scale, universal quantum computers capable of performing any quantum algorithm remain a long-term goal, more targeted processing units already exist. Quantum annealers specialise in solving optimisation problems, whereas quantum simulators are ad hoc devices designed to model specific quantum phenomena (e.g., chemistry) that are hard to study on classical computers.

Quantum emulators run quantum algorithms on classical computers, from laptops to high-performance supercomputers. These approaches allow scientists to test and refine such algorithms without quantum hardware.

Classical supercomputers can emulate quantum systems with up to around 50 qubits, the basic units of information in quantum computing. Despite this limitation, emulation plays a critical role in developing quantum software by enabling researchers to experiment with quantum logic, debug algorithms and explore potential applications in a controlled environment.

AI helps to make quantum computing more efficient and reliable

As algorithms increasingly optimise quantum emulation processes, AI becomes a powerful enabler. For example, AI can help identify the most efficient ways to represent quantum states on classical hardware, reduce the computational overhead of simulating entanglement, and predict the behaviour of quantum systems under different conditions. AI can also help with one of the most critical challenges in large-scale quantum computing: error correction.

Qubits are highly sensitive to ‘noise’, i.e. environmental factors, such as temperature and interference. They lead to frequent calculation errors. Today’s most advanced superconducting quantum chips experience errors in two-qubit operations roughly every 100 to 1,000 operations, corresponding to error rates between 1% and 0.1%.

To achieve meaningful applications in fields like quantum chemistry or materials science, error rates must be reduced by three orders of magnitude to around 0.0001%. To do this, AI techniques are being deployed to manage error correction in two main ways:

• Error mitigation: algorithms monitor quantum circuits in real time, adjusting system parameters to stabilise qubit behaviour and reduce the likelihood of errors. AI can improve the accuracy of qubit measurements by filtering out noise and enhancing signal clarity—crucial for reliably determining quantum states.
• Quantum error correction encodes information across multiple physical qubits to form logical qubits. AI assists in detecting and correcting errors based on observed noise patterns—akin to noise-cancelling headphones.

Reinforcement learning, a branch of AI that learns through trial and feedback, is also being explored to dynamically adapt error correction in response to changing environmental noise conditions. This seeks to allow quantum systems to become more resilient. Over time, they can become self-correcting.  

AI is also helping to design new, more efficient error correction codes tailored to specific quantum hardware architectures, accelerating progress toward fault-tolerant quantum computing.

In some cases, reinforcement learning is being used to tailor quantum circuits to individual devices. Since every quantum computer has unique noise characteristics and operational quirks, reinforcement learning agents can learn to adapt algorithms to the specific behaviour of a given machine. Some companies are pioneering this approach to improve hardware performance.

AI could make quantum computing more accessible

AI is becoming indispensable for making quantum computing more reliable, scalable and commercially viable. It is emerging as an interface between users and quantum hardware, with some companies developing AI-driven “operating systems” that translate natural-language problem statements into quantum algorithms, select the most appropriate combination of quantum and classical computing resources and execute tasks automatically. This approach could make quantum computing accessible to non-specialists and accelerate adoption across industries.

AI can enable quantum networks and communications

Quantum communication is a rapidly advancing field that could enhance data transmission security by leveraging quantum properties to encode and transmit information. Quantum capabilities, such as quantum random number generation (QRNG) and quantum key distribution (QKD), can strengthen encryption and deter attempts to intercept communications, making them highly attractive for cybersecurity. However, building and managing quantum communication networks is technically complex and resource-intensive.

AI could make these networks more practical and scalable. One of the biggest challenges in quantum communication is dealing with errors and signal loss over long distances. AI can help by identifying patterns in transmission errors and applying real-time corrections, thereby improving the reliability of quantum links.

AI could also support resource management and network optimisation. Future quantum networks may involve limited and expensive components, such as quantum repeaters and memory nodes. AI algorithms could dynamically allocate these resources, optimise routing paths and adapt to changing network conditions.

There is a role for AI to play in designing and simulating quantum communication protocols. AI models can suggest more efficient configurations and help researchers test new ideas before deploying them on physical systems. Moreover, because quantum and classical networks operate on fundamentally different principles, they require novel algorithms and intelligent control mechanisms to manage hybrid architectures. AI may play a central role in this orchestration, enabling seamless interoperability by managing routing, resource allocation and protocol translation across both domains.

On the road to the commercialisation of quantum communication solutions, AI could help scale up infrastructure, ensure security and manage complexity. Deploying AI will help orchestrate the integration of quantum and classical networks, an important step for future quantum networks.

Smarter quantum sensing with AI

Quantum sensors can detect extremely subtle changes in physical quantities such as magnetic fields, gravity or luminosity with extreme precision. Their potential applications span areas such as healthcare, navigation and environmental monitoring. However, the data they produce is often complex to interpret.

AI helps tackle this challenge. Machine and deep learning algorithms are increasingly used to filter noise, enhance signal clarity, and even suggest optimal configurations for quantum sensing experiments. This improves measurement quality and reduces the time and cost needed to analyse data and obtain reliable results.

Beyond reducing noise, AI helps turn quantum sensors’ complex data into useful insights. For example, AI could analyse highly detailed brain scans from quantum sensors to spot early signs of neurological conditions. It could also help interpret novel biomedical data, analysing how drugs interact with cells and detecting subtle disease biomarkers.

AI-enhanced quantum navigation is emerging as an alternative to GPS, leveraging quantum sensors to detect minute variations in Earth’s magnetic field. When these sensors are combined with AI algorithms, they can map and interpret geomagnetic signatures with high precision, enabling GPS-independent positioning.

For example, a humanitarian aircraft flying near a conflict zone could find its GPS signals jammed or spoofed, potentially sending the plane off course. With quantum navigation, the crew would no longer rely on satellites. Instead, an aircraft’s sensors pick up magnetic readings along its trajectory, which the AI matches to a global map, and the onboard system instantly calculates the aircraft’s position.

This technology is not limited to the skies. Autonomous underwater drones could navigate deep-sea missions, while ground vehicles could move through tunnels and other areas where satellite signals are unreliable.

Win-win: How quantum can support AI

As quantum technologies continue to evolve, integrating AI will be essential for turning emerging capabilities into real-world applications. A third and final article in this series will flip the perspective: instead of asking how AI can advance quantum technologies, we examine how quantum technologies may ultimately accelerate and elevate AI itself, revealing the full, two-way nature of this emerging partnership.