Quantum technology continues to be either a buzzword or a mystery to many organisations. In a recent SAS study, 60% of business leaders across China, France, Mexico, the UK, and the US said they are actively investing in or exploring opportunities with quantum AI. However, barriers to adoption include cost (38%), lack of understanding or knowledge (35%), and uncertainty around practical, real-world uses (31%).
At the SAS Innovate 2025 conference in Orlando, Florida, Bryan Harris — Chief Technology Officer at SAS — spoke to the media about the challenges surrounding quantum computing and quantum AI, and how those challenges are playing out across industries.
Debunking myths
One of the common misconceptions about quantum AI, Harris noted, is the belief that organisations need to own a quantum computer.
“There are quantum computer providers out there, and you can interact with them just like you would a cloud provider. You can send data, call an API, and have it reason over data with quantum computing. It’s like a consumption-based model for a cloud provider. I think it’s important to let people know that first of all, that barrier is not as high as they think,” he said.
Another issue, he added, is that many people do not understand which types of business problems require a quantum approach, making education crucial.
“Think of a problem that has no algorithm. If I’m solving something with linear regression, I have a set of data points, and I draw a line through them that minimises the error. That line lets me identify a trend and make predictions — but that kind of approach doesn’t work for the problems I’m describing,” Harris explained. “In these cases, you have to try every possible option, and for each one, you need to understand its impact. That’s why it’s critical to help customers determine whether they’re dealing with an optimisation problem. If they are, and if the problem is large or complex enough, it may be a strong candidate for quantum computing.”
Harris added that in optimisation problems, there is often no efficient algorithm, making the process essentially trial and error.
“You’re essentially working with an objective function — something you want to either minimise or maximise. That’s the core of the strategy. Say you have 1,000 possible combinations. You’d need to run each scenario, measure the outcome — maybe it’s profit — and identify which combination gives you the best result. Now imagine scaling that up to a billion combinations,” he said.
In traditional computing, if the number of options is small, conventional solvers can be used, Harris pointed out. But at a massive scale, finding the best solution with traditional CPUs can take six months to a year.
“You may never even get to the best solution because it just takes too long to figure it out,” he noted.
Practical uses
In the consumer goods sector, quantum AI has shown promise in reducing decision-making time. Manufacturing giant P&G needed to identify the right combination of chemicals to manufacture products without cross-contaminating ingredients.
During the first day’s keynote at the conference, Krista Comstock, Director of Digital Innovation at P&G, recalled how they used quantum AI to address this challenge.
The scenario was: P&G had five large tanks where it mixed product ingredients, and needed to determine the optimal assignment of products to tanks to ensure certain products never mixed.
“We’re looking at 100-plus products in five tanks. When you start to think about the scale of that, we’re talking about 10 to the 114th power, which is not small, considering the number of atoms in the universe,” she remarked.
Using traditional computing initially, P&G was able to generate quality and feasible results, but only after six hours.
They then tried a quantum approach. While results were generated in two minutes, the quality and feasibility were lacking.
Eventually, with SAS architecture, P&G adopted a hybrid model combining both traditional and quantum computing. This yielded results in 12 minutes — a 97% reduction in computing time.
“What hybrid quantum architecture does is allow us to search through all the combinations and evaluate them against the objective,” Harris said. “It then provides the top 10 or 20 solutions, which we pass to traditional computing. That reduces the complexity of the search space to a smaller set — maybe 20, 30, or 40 options — and from there, traditional methods can take over. This lets us reach the best outcome at the lowest cost.”
Comstock believes that the hybrid approach is “the best of both worlds,” combining speed with solution quality. As the P&G team moved through the process more quickly, they were able to uncover insights faster and rethink parts of the problem formulation, including new ways of transforming the data.
For quantum AI, SAS’s cloud-native platform SAS Viya enables the hybrid infrastructure to solve what was previously unsolvable, Harris said.
Partner ecosystem
As part of its efforts to make quantum technology more accessible, SAS is focusing on several industry applications — including drug discovery in life sciences, risk modelling in financial services, and process optimisation in manufacturing.
To advance work in quantum AI and analytics, SAS has entered into partnerships with several companies in the space. One is D-Wave Quantum Inc, which develops quantum annealing systems. SAS is using D-Wave’s technology in both its internal research and select customer engagements.
SAS is also a member of the IBM Quantum Network, aiming to accelerate quantum integration through a hybrid approach that combines classical and quantum methods.
In addition, the company is collaborating with QuEra Computing Inc, a developer of neutral-atom quantum computers. Through the QuEra Quantum Alliance Partner Program, SAS is contributing to efforts that aim to broaden the use of neutral-atom systems in solving large-scale computational problems.
“A lot of people say quantum is 10 years from now. No — quantum is right here, right now, and we’re doing it,” Harris said.
