The UK Financial Conduct Authority (FCA) wants firms to test quantum computers on money laundering and fraud detection, the regulator said today (Thursday). Members of its new Quantum Experimentation Group will work on artificial data the FCA has built for the trials.
The group is open to UK financial firms, quantum computing providers and research labs. Registration closes on December 31, and the work starts in the first quarter of 2027.
Data is also where most of the FCA‘s anti-money laundering cases start. Data deficiencies contributed to 68% of its fines in that area over five years, a FinanceMagnates.com review of 22 cases totaling more than £430 million found in January.
London’s trading industry is coming home!
Firms Will Work on Artificial FCA Data
The first use case is quantum machine learning for anti-money laundering and fraud detection. Participants will take a financial crime problem defined by the FCA and compare quantum approaches with classical techniques on the shared synthetic dataset.
The regulator said firms often lack both the specialist skills to turn a problem into a quantum application and suitable data to test it on. The artificially generated dataset is meant to close the second gap, while specialists from quantum providers, research labs and the FCA cover the first.
“We want to help firms explore where quantum could make a meaningful difference,” said Jessica Rusu, chief data, information and intelligence officer at the FCA.
Rusu said the group would look for practical uses for common industry challenges, starting with financial crime. UK fraud losses passed £1.17 billion in 2024, according to UK Finance figures compiled by FM Intelligence.
Quantum projects on the trading side have aimed at other problems. Options Technology, which runs low-latency infrastructure for brokers, opened a quantum computing service in New York in January for portfolio optimization and derivatives risk analytics.
The FCA describes its group as the first initiative of its kind by a regulator. Taking part is not an FCA endorsement of a firm, technology or solution, and participants remain responsible for building and assessing their own models, according to the program page.
A Follow-Up to Two AI Testing Programs
The quantum group follows two AI testing programs the FCA announced in 2025, and in each one the regulator supplies a different input.
AI Live Testing, which Rusu announced in April 2025, lets firms test generative AI tools for consumer-facing services under FCA supervision.
Two months later, the regulator teamed up with NVIDIA on a Supercharged Sandbox that gave early-stage firms accelerated computing and AI software, with testing from October 2025.
Supervision came first and computing power second. In the quantum group, the FCA’s main contribution is its own synthetic dataset.
Encryption Deadlines Run to 2035
The regulator is also treating quantum as a threat. In a blog post yesterday (Wednesday), Mark Francis, director of specialists at the FCA, told firms to start moving to post-quantum cryptography.
A powerful enough quantum computer could break the encryption that protects customer data and transactions today, he wrote.
He pointed to milestones set by the National Cyber Security Centre (NCSC): a migration plan by 2028, high-priority work by 2031 and every system, service and product moved by 2035. Firms should not wait for them, he wrote.
The FCA is trialling a supervisory self-assessment tool with a first sample of firms and plans to consult on cyber resilience guidance that will cover post-quantum risks. Francis noted that failed change projects accounted for about a fifth of all incidents reported to the regulator in 2025.
That consultation will sit next to incident reporting rules the FCA finalized in March, which take effect on March 18, 2027 and cover cyber and third-party failures.
HSBC has already tested a quantum-safe link. The bank used quantum key distribution to protect a EUR 30 million foreign exchange trade in 2023, with BT and Toshiba supplying the network and devices.
Crypto faces the same question. A Google Quantum AI paper argued in March that a machine with fewer than 500,000 physical qubits could break Bitcoin’s elliptic curve encryption in about nine minutes, close to the network’s 10-minute block time.
This article was written by Damian Chmiel at www.financemagnates.com.
