Graphics processing units on today’s computers can only hold so much capacity — and financial institutions are looking to quantum computing to process ever-growing data sets and turbocharge AI. 

“By applying quantum, the parallel [computing] capability of the computer goes up exponentially,” Mitchell Wein, executive principal at Datos Insights, a data and markets analytics company, told Bank Automation News. 

So, what is quantum computing?

At its most basic, quantum computing presents a fundamental change in how information is transmitted and processed, according to Jessica Gonzalez, director of lending strategies at online loan provider Informed IQ. 

Unlike traditional computers that process data in bits, quantum computers process data in qubits — units of quantum information, Wein said. 

A quantum computer can perform parallel processing of data rather than sequential processing, which allows the system to process different sets of data and run different models at the same time, he explained. 

“Basically, this means that I can transmit information using the property of quantum [mechanics] at a much higher speed than I could transmit information using traditional computers,” Wein said. 

Building a quantum computer 

According to a June 2023 report by IBM and UC Berkeley, quantum computers are 100 times faster than traditional computers. In May 2023, IBM entered into a joint deal valued at $100 million over 10 years with University of Tokyo and University of Chicago to develop a quantum-centric supercomputer powered by 100,000 qubits, according to an IBM release. 

In 2022, $2.3 billion was invested in quantum technology startups. Industries like financial services, automotive and life sciences can gain up to $1.3 trillion in value from the technology by 2035, according to an April 2023 McKinsey report.  

While quantum computers can add value to many industries, it is expensive to build one, according to Quantum Zeitgeist, a quantum research and development company. The average research and development cost for a small-scale quantum computer can range from $10 million to $15 million, while the cost of a commercial quantum computer can range from $10 million to $50 million, according to the company. 

Quantum computers are used for solving complex mathematical equations, processing data at a higher speed and running advanced AI algorithms, according to IBM’s website. 

These computers are “able to solve some tasks faster, more accurately and more efficiently than classical computers,” Stefan Woerner, manager of quantum computational science at IBM, said. The speed of AI innovation, too, can benefit from quantum computing, he said.  

Quantum AI 

So, what is quantum AI? 

Coined as recently as 2023, the term quantum AI (QAI) refers to running AI algorithms on quantum computers. This can solve the scalability problem when it comes to understanding data in large language models, Wein said.  

“Quantum AI would just be speeding up what we’re doing with AI: looking at more data elements, processing it and increasing the ability to do it faster,” Informed IQ’s Gonzalez said. 

 As the use of AI in financial services becomes more common, the industry is exploring how quantum computing can be implemented in bank operations to drive security and efficiency, Wein added. 

Institutions looking to utilize quantum computing include $3.4 trillion JPMorgan Chase, $2.9 trillion HSBC and $2.6 trillion BNP Paribas. 

Quantum for secure communications 

BNP Paribas, for one, is preparing its communication security for “Q-Day,” the day when quantum computers will become so efficient that they will be able to decode any kind of encrypted message or information thrown their way, which can be a security threat for sensitive customer data, Anissa Achab, an expert in security and risk management at the bank, told BAN. 

BNPP has 3,600 employees working on cybersecurity and spends 8% to 12% of its IT budget on ensuring the security of its customers’ data and its own data banks, Achab said. Investment in post-quantum cryptography computing is included in this effort, she added. 

BNPP spent 87 million euros ($94 million) on its IT reinforcement costs in the third quarter, according to the bank’s earnings report, and had 185,000 total employees.  

The bank’s post-quantum readiness strategy involves experimenting with new cryptographic algorithms implemented in existing IT systems to transfer consumer data like bank account information and transaction history securely, she said. 

Quantum computing not only improves computing performance but makes data treatment faster, Achab said. That’s a benefit in solving problems too complex for classical computers, which aids in more secure communications. 

For example, today, communications are triggered through satellites, which takes more time because of the distance electrons must travel to reach the satellite — with quantum computers, that travel time is exponentially reduced, Datos Insights’ Wein said. 

Fraud detection 

Other financial institutions are looking to quantum computing for fraud detection. 

In 2023, 91% of fintech, regional banks and community banks in the United States and the United Kingdom reported an increase in fraud attempts, according to a Jan. 25 report based on a query of 450 financial services executives by Alloy, a fraud and risk mitigation company. 

“In fraud detection, one of his biggest problems is false positives, which triggers client dissatisfaction,” IBM’s Woerner said. 

Quantum computing offers solutions for these problems as it gives “a way to get higher accuracy for learning problems with complex structure, [and] detect things like fraud more reliably,” Woerner said. This is because the computer can process a lot of data quickly, he added. 

Meanwhile, Informed.IQ aims to use quantum computing for enhanced real-time fraud detection, Gonzalez said. “This approach leverages the power of qubits in quantum computing, which significantly outperforms traditional binary computing methods in processing speed and complexity handling.” 

JPMorgan also is using quantum computing for fraud detection, according to its website, which notes that the bank has built an internal team of scientists to work on the technology at its Global Technology Applied Research Center in New York. 

JPMorgan has developed quantum algorithms for risk analysis, deep hedging, option pricing and portfolio optimization and recently led a $300 million investment round in Quantinuum, a quantum computer development company, according to a Jan. 16 release. 

“Financial services has been identified as one of the first industries that will benefit from quantum technologies,” Lori Beer, global chief information officer at JPMorgan Chase, said in the release, noting that the investment will help the bank develop the tech for use cases within the bank and the industry at large. 

Quantum computing enables machine learning and AI models to be trained more efficiently, according to a March 2023 study by JPMorgan and quantum computing development company QC Ware. 

Asset management  

HSBC is using quantum computing to explore asset management. 

“The quantum revolution is not on the horizon; it’s already here,” CEO Colin Bell said in a Dec. 6 release that announced the bank had started using quantum computing for foreign exchange trading. The technology was developed with BT, Toshiba and Amazon Web Services. 

Like BNPP, HSBC is using quantum computing’s cryptography to secure trading data against potential cyber threats, the release said. 

Portfolio managers improve diversification of assets, rebalance investments to respond to market conditions more precisely and cost-effectively streamline trading settlement processes, according to IBM’s website. 

Quantum computers can also assist asset managers to maximize returns while balancing risks and remaining compliant in a short amount of time, IBM’s Woerner said, noting that such solutions are called “dynamic asset allocation,” which takes multiple factors into account while investing, quickly and cost-efficiently. 

“This is a very complicated optimization problem, which is very difficult to solve classically” with a traditional computer, Woerner said. 

Insurance outlooks  

Outside of banks, quantum computing also has the potential to revolutionize the way insurance companies assess and price risk, Wein said.  

With their ability to process vast quantities of data and perform complex calculations, quantum computers can help insurers analyze factors such as customer demographics and health records, and develop more accurate and precise actuarial models, Wein said. 

With a traditional computer, “you can only run so many scenarios [simulations],” which is quite resource extensive, meanwhile, a quantum computer “could run all of those scenarios concurrently,” Wein said. 

A quantum computer can go from 10,000 simulations to 10 million or even 10 trillion scenarios with better accuracy in almost the same time frame, making operations more efficient, he said. 

Hurdles ahead 

While quantum computing can increase efficiencies within insurance, asset management, fraud detection and communications, there are still tech obstacles to overcome, including the ability to access the infrastructure and maintain compliance, Informed IQ’s Gonzalez said.  

Leveraging quantum computers for identified uses is less helpful if the rest of the industry is outdated, and this can create hurdles for the rollout and adoption of the technology —regardless of how sophisticated the algorithms are, she said. 

“There’s a lot of opportunities to really make sure that [quantum computing] is being able to be used to its fullest,” she said. 

This is the final week to register for an early-bird discount to attend Bank Automation Summit U.S. 2024 in Nashville, Tenn., on March 18-19! Discover the latest advancements in AI and automation in banking. Register now.