IBM and Lloyds Banking Group have conducted an experiment to see if quantum computing can be used to identify money mules, and the project highlights significant long-term potential.
From fraud detection to optimization and simulation tasks, the computations required in financial services are constantly increasing in complexity. While AI and classical machine learning have an important role to play, Lloyds believes they will eventually reach their limits.
Early research by the company’s Emerging Technology & Innovation (ET&I) team suggests several areas where quantum computing has the potential to outperform traditional methods, including graph-based anomaly detection, a key element in detecting fraud and criminal activity.
“Preventing economic crime, and in particular detecting mule accounts, requires the analysis of highly complex financial transaction networks,” explained Jamie Harbor, Enterprise Architect for Emerging Technologies and Innovation, and Adam Milner, Lead Quantum Ambassador at Lloyd’s.
“These can be represented as graphs of customers, accounts, and payments, but suspicious activity is often hidden in subtle network structures.”
Traditional computers struggle with certain types of graphing problems because the number of possible solutions increases exponentially with the size of the problem. This could potentially be handled more efficiently by quantum computing.
“Our experiment was not designed to explore ways to replace machine learning models currently used for fraud and crime prevention,” Haber and Milner said.
“Instead, we considered whether quantum-enhanced techniques could generate more sophisticated graph-based features to support future models, features that might be too complex or expensive to compute classically.”
Quantum shows long-term promise
The nine-month project focused on graph-based analysis of mule activity using anonymized real-world transaction data on IBM’s cloud quantum computer, and considered several different quantum algorithm approaches.
The aim was not to provide production-ready solutions, but to figure out which quantum technologies really hold promise in the long term.
Some quantum algorithms have proven effective, especially when used to generate new types of features or enable deeper network analysis to complement AI and ML, according to Lloyds.
Alongside the experiment, the ET&I team has also developed a broader roadmap identifying several potential quantum use cases across Lloyds Group.
Some tasks, such as optimization tasks, can become realistic propositions relatively quickly because the algorithms and hardware involved are mature.
“One of the most valuable outcomes of the experiment is capacity building,” Haber and Milner said.
“This experiment has created a hands-on learning opportunity for our colleagues through hands-on code reviews, detailed walkthroughs of algorithmic decisions, and the establishment of a Quantum Ambassadors program, a group responsible for deepening our expertise, exploring new use cases, and supporting the growth of a thriving internal quantum community.”
Follow us on social media
Follow ITPro on Google News and Add us as a preferred source Stay tuned for all the latest news, analysis, views and reviews.
Also, Follow ITPro on LinkedIn, ×, facebookand blue sky.
