AI has taken centre stage over the past few years, securing nearly a quarter of all VC investments in early-stage deals in Europe in 2024— and it’s giving a boost to a lesser-known technology: quantum computing.
Quantum computing businesses are riding the wave of AI by exploring how both technologies could work together. As it turns out, quantum technologies — which are expected to eventually enable calculations that are impossible to carry out with ordinary computers — could unlock significant capabilities for AI.
In Europe, investors say the majority of quantum computing startups are currently exploring how to enable “quantum AI”. Olivier Tonneau, partner at quantum-focused VC firm Quantonation, says that he is seeing this across portfolio companies, and that it is now a must-have for businesses in the space.
“It could be a huge market,” he says. “If you can show that thanks to a quantum computer you can enable a few percentages of efficiency gains for a company like OpenAI, the potential revenue and value creation are enormous.”
What can quantum computers do for AI?
Quantum computers are built on a different paradigm to ordinary computers. Instead of storing information in the form of classical bits, which present as either a “zero” or “one”, quantum computers are built with qubits, which can be both zero and one simultaneously.
This is possible because quantum computers rely on quantum mechanics — complex laws of physics that can be observed at the atomic and subatomic levels. In principle this means that they excel at representing complex physical phenomena that happen at the quantum level, such as interactions between molecules.
Many use cases for AI models rely on simulating these phenomena: understanding and predicting molecular behaviour is central, for instance, to designing better drugs and new materials.
But the classical computers that AI models run on for these use cases need huge amounts of data to simulate the complexity of quantum phenomena, and can only do so with limited precision. Quantum companies are therefore exploring how AI models could leverage quantum computers to enable more accurate and precise results, while requiring smaller datasets.
“Quantum data is the representation of the actual reaction in quantum states. It gives the AI model better and more accurate data,” says Raj Hazra, the CEO of UK-based quantum scaleup Quantinuum.
“Quantum can enable classical AI to do things it couldn’t do before.”
The AI opportunity
Earlier this month, Quantinuum launched a ‘Generative Quantum AI framework’, which it says already enables users to leverage quantum-generated data for AI models. Hazra says that the company’s latest quantum computer, which it launched in 2024, is powerful enough to represent “simpler” quantum effects that are impossible to simulate on ordinary computers.
The company is working with partners in automotive, pharmaceuticals and materials science, with the results of quantum AI experiments carried out in chemistry and finance soon to be published, says Hazra.
Hazra declined to specify how much revenue Quantinuum expects to generate. Commercialisation will ramp up as the company’s quantum computers scale and become capable of representing more complex phenomena in the coming years, he tells Sifted.
The sector has been held back by the capabilities of quantum hardware. Controlling qubits, which are highly unstable, is a major barrier to scaling — meaning that quantum computers are still limited in size and capabilities.
Loïc Henriet, the CEO of French quantum computing startup Pasqal, says that although the company is seeing proofs of concept emerge in which quantum AI models have shown better performance on small-scale problems, these remain “academic problems” rather than real-life use cases.
“There are still bolts to unlatch before we have something that is competitive and in production,” he says.
Despite the potential, therefore, it is still early days. “We don’t currently have a powerful enough quantum computer to be able to train AI models,” says Quantonation’s Tonneau. “These are two sectors that are going to converge strongly and work together, but experiments are still in their infancy.”
Investing in quantum computing
Pasqal is not yet driving “significant” revenue from quantum AI, but it remains “an important part” of the company’s activity, says Henriet.
The startup has partnered with hardware providers like Nvidia to explore where quantum processors can fit in an AI workflow, while also working with software companies such as Qubit Pharmaceuticals, which leverages the technology for drug discovery, building domain-specific algorithms.
Most other startups in the space, from France’s Quandela and the UK’s Oxford Quantum Circuits to Finland’s IQM, are exploring the application of quantum computing to AI — and quantum computing’s largest players like IBM are also dedicating R&D teams to the topic.
“It’s a perspective that is gaining traction among all players,” says Henriet.
As a growing number of companies integrate AI into their workflows and unlock larger budgets to maximise the technology’s potential, a market opportunity for quantum computing is emerging, adds Henriet.
“It’s seen by quantum providers and by investors as a strong opportunity to join a market that is growing very fast,” he says.
Sifted data shows that funding for quantum technologies across Europe reached nearly €730m in 2024 — a way off the €10.5bn that were pumped into AI startups in the region that same year, according to Atomico.
“The fact that there are applications in AI is raising interest in quantum,” says Tonneau. “But there is still a long way to go.”
“Investment in quantum is still anecdotal compared to AI.”
Read the orginal article: https://sifted.eu/articles/quantum-computing-ai/
