How Quantum Computing Could Improve Generative AI: What a New Drug-Discovery Study Reveals

By Mohsen Rahmani, Matthew Woolway, Vladimir Vargas-Calderón, William Kim, Kevin Chern, and Mohammad Amin

A newly published, peer-reviewed study in Scientific Reports offers some of the clearest evidence yet that annealing quantum computers could meaningfully improve the outputs of generative AI models.  

The research, titled “Molecular design beyond training data with novel extended objective functionals of generative AI models driven by quantum annealing computer,” was conducted in collaboration between D-Wave and Shionogi & Co., Ltd. (“Shionogi”) (formerly the pharmaceutical division of Japan Tobacco). Together, we used a D-Wave™ annealing quantum computer to improve how an AI model designs new drug candidate molecules.  

The result: molecules that were more chemically valid and more “drug-like” than those produced by an equivalent classical-only AI model. The molecules generated by the quantum-assisted model even exceeded the quality of the very data the model was trained on.  

These results matter because drug discovery is notoriously slow, expensive, and prone to failure. Bringing a single new drug to market typically takes 10 to 15 years with a median cost of $708 million, according to a 2025 RAND Corporation study. Much of that expense and time is spent synthesizing and testing compounds that ultimately don’t pan out. Technologies like quantum AI that can narrow the field earlier—by generating a shorter list of candidates that are more likely to be valid, drug-like, and worth pursuing in the lab—attack the problem at its most expensive stage.

Why Does Generative AI Struggle to Design Good Drugs?

Generative AI has shown promise to be a powerful tool for the early stages of drug discovery. Instead of screening existing compound libraries, a model can generate entirely new molecular structures. But there's a persistent challenge of generating molecules that meet all four essential criteria required for useful drugs: valid chemistry, desired activity against the target proteins, acceptable chemical properties for use in the body (low toxicity, stability, easy absorption, etc.), and the ability to be easily synthesized. Reliably hitting all four criteria simultaneously is still the field's central bottleneck.

The space of chemically plausible, drug-sized molecules is estimated at more than 10⁶⁰, while the number of compounds humans can currently synthesize and test is closer to 10¹⁰. Generative models trained on the relatively tiny, synthesizable slice of that space tend to overfit, producing outputs that are either invalid chemistry or poor drug candidates. Researchers call this the drug-likeness problem, and it’s one of the main reasons AI-generated molecules still need heavy filtering before any candidates can be synthesized.

How D-Wave and Shionogi Combined Classical AI with Annealing Quantum Computing for Drug Discovery 

To find out whether annealing quantum computing could help, we built a generative model—a transformer-based model augmented with a Boltzmann machine—and tested two versions. One ran entirely on classical hardware, and one used the D-Wave annealing quantum computer as part of its sampling process. Comparing the two head-to-head, on the same data and architecture, isolated the contribution of D-Wave's annealing quantum hardware. 

One of our key inventions that made this pairing possible was a neural hash function (NHF), which converts the model's internal representation into the binary code the quantum hardware operates on, while also acting as a regularizer that keeps training stable, replacing the Gumbel-Softmax binarization technique used in discrete variational autoencoders. It's this custom-built bridge that let the classical network and the annealing quantum system train together as a single, coherent model rather than two loosely connected systems. 

That binary code feeds into a Boltzmann-machine prior, and this is where the D-Wave hardware does its work. Rather than approximating that prior with classical Markov chain (or similar) Monte Carlo methods, we sampled it directly on a D-Wave Advantage2™ quantum annealing system, drawing samples closer to the model’s true target distribution.  

In effect, the annealing quantum component serves as the model’s sampling engine, a role classical computers can only approximate.

Can Annealing Quantum Computing Improve Generative AI Outputs? 

Yes, annealing quantum computing can improve generative AI outputs. In the drug discovery study, we demonstrated that the model using annealing quantum computing produced chemically valid molecules 97% of the time, compared with 73% for the same architecture running on a classical Boltzmann machine, and 54% for a previously published benchmark model on the same dataset.  

On drug-likeness—measured by the Quantitative Estimate of Drug-Likeness (QED) score, a standard metric used to measure a compound’s favorability to become an oral drug—the results were just as striking. The study defines “drug-like” molecules as those scoring QED ≥ 0.7, and reports what share of unique generated molecules met that bar: 31.61% for the original training data, 43.15% for the best classical-only model, and 66.79% for the annealing quantum model.  

In other words, the annealing quantum model generated drug-like molecules at more than double the rate present in the training data and about 55% higher than the equivalent classical-only model. 

What Does This Study Mean for AI Beyond Drug Discovery? 

The implications of this research for AI are not limited to drug discovery. Based on the results, annealing quantum computing could act as a general-purpose stochastic generator inside generative AI, sampling more effectively from the “in-between” spaces that a model hasn’t directly seen in training. This is exactly the capability that limits generative AI across many domains, not just molecular design.  

This isn't a hypothetical claim about quantum computing's future value. It's consistent with what D-Wave's annealing quantum computers are already doing in customer applications, addressing real-world optimization problems today across manufacturing, telecommunications, retail, logistics, and defense. In each of these customer applications, the underlying task is the same: efficiently searching or sampling a large combinatorial space to find good solutions faster than classical methods alone. 

The study with Shionogi extends that same core capability—efficient sampling over vast, complex spaces—from operational optimization problems into the training and sampling process of a generative AI model itself. That's a meaningful expansion of where annealing quantum computing can add value.

Key Takeaway 

This is a peer-reviewed, real-world demonstration that an annealing quantum computer can improve the output quality of a generative AI model on a task that matters commercially: drug discovery.  

For an industry used to hearing about the promises of quantum computing’s future potential, this study is notable for measuring something concrete: validity rates, drug-likeness scores, and a direct comparison against classical hardware running the identical model. These results show promising evidence that annealing quantum computing can act as a practical enhancement to AI today, not just a long-term bet. 

The full study, “Molecular design beyond training data with novel extended objective functionals of generative AI models driven by quantum annealing computer,” is published open-access in Scientific Reports (July 2026) and available at nature.com.