Federal quantum conversations often start with cryptography: What happens when a powerful enough quantum computer can break today’s public-key encryption? Bill Wisotsky, principal quantum systems architect at global data and software company SAS, says agencies should also look at quantum’s potential role in optimization, artificial intelligence (AI), machine learning, and infrastructure planning.
In a recent interview with MeriTalk, Wisotsky said quantum’s value for government will likely show up in logistics, scientific discovery, and other areas where agencies face complex computational problems.
“Quantum has the ability to change the world,” Wisotsky said.
For Wisotsky, the federal opportunity starts with a practical mindset: Quantum should be treated as another accelerator inside a hybrid technology environment, not a wholesale replacement for classical computing. Agencies that begin with well-defined pain points, test low-risk pilots, and build trusted partnerships will be better positioned when the hardware matures.
“If you wait until fault-tolerant, you’re way far behind,” he said.
Quantum’s growing federal role
Wisotsky first encountered quantum computing in graduate school in the late 1990s and early 2000s, when it still felt largely theoretical. The recognition that quantum could one day threaten core encryption schemes moved the technology from academic theory into national security and industry conversations.
“If a foreign government could break public-key encryption, we are doomed,” he said. “That threat kind of launched quantum onto the main stage, not just in government, but in industry as well.”
In recent years, the government conversation has widened. Wisotsky said agencies and policymakers increasingly understand that quantum could help with optimization, AI, and machine learning. That has placed quantum in a strategic technology race, similar to the competition around outer space and semiconductors.
He also noted that federal attention to quantum has remained relatively steady across administrations, pointing to the National Quantum Initiative Act as an example of continuity.
“The importance of it seems to be understood across the board,” Wisotsky said.
The optimization opportunity
One of the clearest government use cases is optimization, Wisotsky said.
In a previous role with SAS Federal, Wisotsky worked with defense and homeland security customers and saw firsthand how many government operations depend on complex scheduling, routing, resource allocation, and planning scenarios.
“So much of the government is optimization problems,” he said.
Those problems aren’t just about vast data volumes. More importantly, they are about relationships within the data. When agencies confront vast combinations of variables, quantum computing offers a way through the noise, Wisotsky explained.
“If the solution space is narrow, you could do that classically pretty easily,” he said. “But if the solution space is very, very broad with a lot of different possibilities, that’s where quantum really comes in handy.”
Wisotsky pointed to logistics, supply chains, scheduling, disaster response, traffic flow, infrastructure, and satellite placement as examples of areas where agencies should explore quantum’s potential.
But he cautioned that optimization problems do not automatically call for quantum. Quantum systems are probabilistic, while many government problems require confidence that the answer is the best available answer, not just a good one.
That is why Wisotsky emphasized hybrid quantum-classical workflows. He described one approach where quantum-generated solutions are fed into a classical optimizer as a “warm start,” giving the classical system a better starting point to prove optimality faster.
“I think that hybrid approach is really important,” he said.
Early quantum pilots
Some federal quantum efforts appear appropriately focused on long-term, fault-tolerant systems, Wisotsky noted. But he warned that waiting for that milestone before experimenting could leave agencies behind.
Instead, agencies should run small pilots and proofs of concept, especially around hybrid processes for machine learning, AI, optimization, and related mission problems. Those pilots can expand as quantum hardware becomes more powerful, he advised.
Quantum-inspired thinking is a near-term benefit of quantum experimentation, Wisotsky said. In one example, he said SAS explored quantum support vector machines and, in the process, found a faster way to reformulate a classical problem without using quantum hardware.
“We realized if we thought about the classical problem differently, we could reformulate it, and we created a new classical algorithm that was 50 times faster than the one that we were using,” Wisotsky said.
Value beyond hype
Wisotsky said quantum’s hype problem cuts two ways: Some claims overstate what the technology can do today, and others miss benefits that have little to do with raw speed.
Exaggerated online claims, including examples suggesting that quantum computers can solve artificial benchmark problems in minutes that would take classical computers millions of years, can mislead government and industry leaders, he said. Often, those problems are designed specifically to showcase quantum behavior rather than solve real-world mission needs, he noted.
“Those types of things really do harm for quantum,” he said.
Wisotsky also pushed back on two common misconceptions: that quantum is always dramatically faster, and that it is especially useful for big data. In some AI and machine learning work, Wisotsky said SAS has found quantum approaches are slower than classical methods.
But “advantages don’t always have to do with speed,” he observed.
In some cases, Wisotsky said, quantum machine learning may be able to achieve similar accuracy with far less data. That can benefit agencies that cannot easily store, label, move, or train on massive datasets. Quantum systems may also produce models that are more expressive, require less data, and potentially remain stable longer as data changes over time, he noted.
Wisotsky used the analogy of adding dye to a fish tank: A drop of blue dye noticeably changes a small tank, but the same drop is absorbed almost invisibly in a much larger one. In a similar way, a higher-dimensional quantum representation may be less disrupted by small changes in the data, he suggested.
Wisotsky also pointed to energy consumption as a potential advantage, particularly as AI workloads raise concerns about power demand.
“Quantum computers use far less power than high-performance computing centers,” he said.
Mission-driven adoption
For federal chief information officers, chief information security officers, and mission leaders, Wisotsky’s advice was to begin with the problem, not the technology.
The first question is whether a problem can be solved classically. If it can, and if a classical approach delivers better results, that may be the right answer. If part of the problem remains difficult, quantum may have a role.
As government buyers evaluate vendors and other industry partners, Wisotsky advised caution about working with companies whose business models depend on selling quantum time regardless of whether quantum is needed.
“Every problem can be solved with quantum, even if it doesn’t need to be,” he said.
Instead, he said agencies should identify specific pain points, ask potential partners why current approaches are falling short, build relationships across industry, and run low-risk proofs of concept.
Streamlined management
At SAS, leaders view quantum as another computing accelerator alongside central processing units (CPUs), graphics processing units (GPUs), and quantum processing units (QPUs).
“Our goal is to provide value to our customers, whether it be on CPUs, GPUs, or QPUs,” Wisotsky said.
SAS’s 50-year history as an analytics software firm shapes its approach to quantum opportunities, he noted. The company brings industry experts and classical algorithms to bear first, asking whether a problem can be solved through existing methods, whether the formulation needs to change, or whether it is truly a quantum candidate.
“We wouldn’t approach a government contract and say, ‘Yeah, let’s just do quantum,’” he said. “We would really examine the problem, try to formulate a solution with our experts, and ask, ‘Do we need to use quantum and, if we do, how?’”
SAS recently launched a quantum lab that Wisotsky said is designed to lower the barrier for users managing multiple quantum environments, software development kits, and hardware-specific workflows.
The goal, he explained, is to let users build quantum circuits in the familiar SAS language while SAS handles more of the back-end complexity. Over time, SAS wants to bring quantum models into the same kinds of environments where users already compare and manage classical models.
Next steps for quantum readiness
Wisotsky said quantum today reminds him of earlier technology shifts that looked awkward at first, such as cassette-based computers, suitcase-sized cell phones, and dial-up modems. Each seemed limited before it became part of something much larger.
Quantum may be at a similar point, he said. Better hardware will matter, but so will a different way of thinking.
Rather than lifting classical algorithms into quantum form, Wisotsky said researchers and agencies may need to rethink problems from the ground up.
“When we stop thinking about how we do things now, that’s when we’re going to really see a tremendous benefit,” he said.
For agencies, the practical move is to start small: identify hard optimization problems, test hybrid approaches, and build the expertise to separate useful quantum applications from vendor claims.
“The government has a lot to gain,” Wisotsky said.