Quantum Algorithms: A New Frontier in Financial Crime

Prevention
Abraham Itzhak Weinberg1 and Alessio Faccia2
arXiv:2403.18322v1 [cs.LG] 27 Mar 2024

1
AI-WEINBERG, AI Experts, Tel Aviv, Israel, [email protected]
2
University of Birmingham Dubai, Dubai, UAE, [email protected]

March 28, 2024

Abstract
Financial crimes’ fast proliferation and sophistication require novel approaches that provide
robust and effective solutions. This paper explores the potential of quantum algorithms in
combating financial crimes. It highlights the advantages of quantum computing by examining
traditional and Machine Learning (ML) techniques alongside quantum approaches. The study
showcases advanced methodologies such as Quantum Machine Learning (QML) and Quantum
Artificial Intelligence (QAI) as powerful solutions for detecting and preventing financial crimes,
including money laundering, financial crime detection, cryptocurrency attacks, and market
manipulation. These quantum approaches leverage the inherent computational capabilities
of quantum computers to overcome limitations faced by classical methods. Furthermore, the
paper illustrates how quantum computing can support enhanced financial risk management
analysis. Financial institutions can improve their ability to identify and mitigate risks, leading
to more robust risk management strategies by exploiting the quantum advantage. This research
underscores the transformative impact of quantum algorithms on financial risk management.
By embracing quantum technologies, organisations can enhance their capabilities to combat
evolving threats and ensure the integrity and stability of financial systems.
Keywords: Quantum Finance, Financial Crimes Mitigation, Artificial Intelligence, Quantum
Algorithms, Quantum Machine Learning (QML), Machine Learning (ML), Cryptocurrency

1 Introduction
Entering the Quantum Decade, as highlighted by IBM [1], signifies the beginning of a transforma-
tive period in quantum computing. This decade is expected to see the development of quantum
computing from theoretical exploration to practical applications that tackle some of the most com-
plex challenges across various industries. IBM’s Quantum Network, a global community of over
210 Fortune 500 companies, leading academic institutions, startups, and national research labs,
illustrates a collective effort to utilise the potential of quantum computing. This network exem-
plifies collaborative efforts to accelerate progress in quantum computing, with the ultimate aim of
developing applications that solve previously unsolvable problems. As we explore the capabilities
of quantum computing, especially in improving quantum algorithms for fighting financial crimes,

1
it is clear that the Quantum Decade is not just a prediction of technological progress but a call to
action for organisations to prepare for the quantum revolution.
Quantum Machine Learning (QML) strives to develop and implement techniques that can be exe-
cuted on quantum computers to tackle the standard tasks of supervised, unsupervised, and Rein-
forcement Learning (RL) found in classical Machine Learning (ML). QML’s utilisation of quantum
operations sets it apart, coupling the extraordinary properties of quantum computing, such as su-
perposition, tunnelling, entanglement, and quantum parallelism. This paper relates to Quantum
Neural Networks (QNNs), the quantum equivalent of classical neural networks.
QML represents a complex intersection of quantum computing and machine learning (ML), aiming
to revolutionise how we approach data analysis, pattern recognition, and decision-making processes.
QML addresses the core tasks of supervised, unsupervised, and reinforcement learning central to
classical ML but with a distinctive twist by leveraging techniques executable on quantum computers.
Integrating quantum operations—taking advantage of superposition, tunnelling, entanglement, and
quantum parallelism—enables QML to process information in ways that classical systems cannot
match. At the heart of this innovative field are Quantum Neural Networks (QNNs), which embody
the quantum counterpart to classical neural networks. QNNs promise to enhance computational
capabilities significantly [2], offering new avenues for complex problem-solving and algorithmic ef-
ficiency. Unlike their classical counterparts, QNNs can theoretically handle vast datasets with
unparalleled speed, thanks to their ability to operate in high-dimensional Hilbert spaces. This
capability is particularly promising for tasks requiring large-scale data analysis in bioinformatics,
climate modelling, and financial forecasting. Furthermore, QML is not merely about transposing
existing ML algorithms into the quantum domain; it includes creating new algorithms native to
quantum computing. These quantum-specific algorithms are designed to exploit the natural advan-
tages of quantum states, potentially leading to algorithmic complexity and computational speed
breakthroughs. For instance, quantum algorithms for pattern recognition can operate more effi-
ciently by exploring all possible patterns simultaneously due to quantum superposition.
Additionally, QML’s exploration of quantum entanglement offers novel ways to represent and un-
derstand complex correlations in data that classical algorithms struggle to untie [3]. This aspect
is particularly intriguing for developing more sophisticated models of human cognition and social
networks, where the intricacies of interconnections and influences are beyond the reach of classical
computation. The synergy between quantum computing and machine learning, as embodied in
QML and QNNs, heralds a new era of computational intelligence. By transcending the limitations
of traditional computing paradigms, QML opens the door to solving some of the most challenging
problems facing various scientific and industrial fields today. As we stand on the brink of this
computational revolution, the potential applications and implications of QML and QNNs continue
to unfold, promising profound impacts on technology, society, and our understanding of systemic
complexity.
Table 1 below categorises the most relevant types of Financial Crimes. It offers a comprehensive
framework for classifying financial crimes, integrating and harmonising key concepts from globally
recognised entities such as the Association of Certified Fraud Examiners (ACFE), the Financial
Action Task Force (FATF), Anti-Money Laundering (AML) standards, the United Nations Office
on Drugs and Crime (UNODC), and the International Monetary Fund (IMF). This framework en-
capsulates the vast range of criminal activities these organisations aim to combat by categorising
financial crimes into broad areas such as fraud and scams, money laundering, terrorist financing,
corruption and bribery, cybercrime, market abuse, tax evasion, and identity theft. Each category
and subcategory reflects the definitions and concerns highlighted by these entities, ensuring a broad

2
coverage that includes various types of financial crimes identified in their reports and recommen-
dations.
The framework’s inclusivity allows for a structured approach to understanding financial crimes,
offering a solid foundation for further analysis, policy formulation, and the development of coun-
termeasures. For instance, the FATF’s focus on money laundering and terrorist financing finds a
direct correspondence in Table 2, ensuring that the global priorities in combating financial crimes
are well-represented. Similarly, the ACFE’s extensive work on fraud and corruption is mirrored
in the detailed subcategories under ‘Fraud and Scams’ and ‘Corruption and Bribery’, respecting
the depth of their research and guidance. The inclusion of cybercrime acknowledges the growing
concern of entities like the UNODC and IMF regarding the digital dimension of financial crimes,
emphasising the evolving nature of these threats. Moreover, the framework’s structure facilitates
easy mapping and integration of specific guidelines, recommendations, and reporting requirements
from these organisations, making it a versatile tool for academic and practical applications. It
supports the alignment of national and international efforts in the fight against financial crimes,
encouraging a unified approach that can enhance effectiveness and cooperation across borders.
In essence, Table 1 serves as a foundational piece in the larger puzzle of global financial crime pre-
vention, offering a starting point to explore, match, and integrate the diverse and rich perspectives
of leading organisations in the field. It aims to foster a holistic understanding and approach to
combating financial crimes, reflecting the complexity and interconnectivity of these illegal activities
in the modern world.

3
Category Subcategory Description
Fraud and ”Scams” Consumer Fraud Deceptive practices causing loss in con-
sumer transactions.
Corporate Fraud Misconduct within corporations for finan-
cial gain.
Securities and Investment Market manipulation, Ponzi schemes, in-
Fraud sider trading.
Insurance Fraud False claims to obtain insurance benefits.
Money Laundering and Traditional Money Launder- Concealing origins of illegally obtained
Proceeds of Crime ing money through transactions.
Cryptocurrency Laundering Using digital currencies to hide the source
of funds.
Asset Laundering Reinvesting criminal proceeds into assets
to conceal origins.
Terrorist Financing Domestic Financing Funding terrorist activities within a coun-
try.
International Financing Transferring funds across borders to sup-
port terrorism.
Corruption and Bribery Public Sector Corruption Misuse of official powers in the public sec-
tor for personal gain.
Private Sector Corruption Bribery and misconduct in private indus-
try.
Transnational Bribery Bribes paid to foreign officials for business
advantages.
Cybercrime Financial Data Breaches Unauthorized access to financial data lead-
ing to theft or fraud.
Online Fraud and Scams Cyber methods targeting individuals or or-
ganizations for financial gain.
Ransomware and Extortion Demanding payment to prevent or undo
cyberattacks.
Market Abuse and In- Market Manipulation Artificially inflating or deflating the price
sider Trading of securities.
Insider Trading Trading based on non-public, material in-
formation.
Tax Evasion and Avoid- Domestic Tax Evasion Illegally hiding income from national tax
ance authorities.
Offshore Tax Evasion Using foreign jurisdictions to evade taxes.
Aggressive Tax Planning Exploiting loopholes to minimize taxes
without direct illegality.
Identity Theft and Fi- Identity Theft Using someone else’s personal information
nancial Impersonation for financial gain.
Financial Impersonation Assuming a false identity to commit finan-
cial crimes.

Table 1: Categories of Financial Crimes.

4
Financial Crimes Technological ML Countermea- Examples of Quantum Algo-
Countermeasures sure Approach rithms for Countermeasure Ap-
proaches
Consumer Fraud AI-powered fraud de- Supervised learning for Quantum anomaly detection algo-
tection systems pattern recognition rithms use Quantum Support Vector
Machines (QSVM) to identify un-
usual patterns.
Corporate Fraud Digital forensic tools Network analysis for Grover’s search algorithm for fast
detecting fraudulent retrieval enhances quantum cluster-
transactions ing algorithms to identify groups
with similar fraudulent patterns.
Securities and Invest- Blockchain for trans- Anomaly detection al- Quantum Principal Component
ment Fraud parent transactions gorithms Analysis (QPCA) for dimension-
ality reduction in large datasets,
improving classification accuracy.
Insurance Fraud Predictive analytics for Classification algo- Quantum algorithms for pattern
claim analysis rithms for fraudulent recognition and anomaly detection
claims in claims data, potentially using
Quantum Approximate Optimisa-
tion Algorithm (QAOA).
Traditional Money AI-powered transac- Clustering algorithms Quantum machine learning (QML)
Laundering tion monitoring for detecting suspi- for network analysis, using
cious activities quantum-enhanced clustering
to reveal hidden structures in
transaction data.
Cryptocurrency Laun- Blockchain analytics Anomaly detection in Quantum simulations to trace com-
dering tools cryptocurrency trans- plex cryptocurrency flows and iden-
actions tify laundering schemes with un-
precedented speed and accuracy.
Asset Laundering Digital asset tracking Machine learning mod- Quantum-enhanced algorithms for
systems els for asset tracing analysing asset registration and
transaction patterns to uncover il-
licit financial flows.
Domestic Financing Network analysis tools Regression analysis for Quantum algorithms for complex
fund tracking network simulations, aiding in the
identification of covert funding
channels for terrorist activities.
International Financ- Cross-border transac- Deep learning for pat- Quantum simulations to model and
ing tion monitoring tern recognition in fi- analyse international financial net-
nancial flows works for tracking terrorist financ-
ing across borders.
Public Sector Corrup- E-governance plat- Text mining for Quantum computing techniques for
tion forms anomaly detection in deep analysis of government pro-
procurement data curement data to identify patterns
indicative of corruption.

5
Private Sector Corrup- Compliance monitor- Predictive analytics for Quantum-enhanced pattern recogni-
tion ing AI risk assessment tion in corporate transactions and
communications to detect bribery
and misconduct.
Transnational Bribery International database Machine learning for Quantum algorithms for analysing
integration identifying risk pat- global financial transactions to un-
terns cover transnational bribery schemes.
Financial Data Encryption technolo- Anomaly detection for Quantum key distribution (QKD)
Breaches gies unauthorised access for secure communications, ensuring
data integrity against cyber threats.
Online Fraud and Behavioural analytics Supervised learning for Quantum-enhanced machine learn-
Scams email filtering ing for real-time fraud detection, sig-
nificantly reducing false positives in
online transactions.
Ransomware and Ex- Anti-malware tools Predictive models for Quantum-resistant encryption
tortion with AI threat detection methods to safeguard data against
ransomware attacks and ensure
privacy.
Market Manipulation Real-time market Time series analysis Quantum optimisation algorithms
surveillance systems for detecting manipu- to solve complex market prediction
lation models more efficiently than classi-
cal algorithms.
Insider Trading Anomaly detection in Classification models Quantum algorithms for analysing
trade data for unusual trading vast amounts of market data and
patterns communications to detect insider
trading activities.
Domestic Tax Evasion Automated tax report- Regression models for Quantum Fourier Transform (QFT)
ing and analysis sys- income prediction algorithms identify cyclical patterns
tems in financial data indicative of eva-
sion schemes.
Offshore Tax Evasion International tax com- Network analysis for Quantum-enhanced analysis of
pliance databases offshore entities global financial transfers to detect
and trace offshore tax evasion
strategies.
Aggressive Tax Plan- AI for legal document Machine learning for Quantum computing is used to anal-
ning analysis loophole identification yse complex tax planning strate-
gies and identify potentially abusive
schemes.
Identity Theft Biometric verification Machine learning for Quantum-enhanced machine learn-
systems biometric authentica- ing for real-time identity verifica-
tion tion, significantly reducing the risk
of identity theft.

6
 Financial Imperson- Digital identity verifi- Deep learning for face
Quantum simulations for detecting
ation cation platforms recognition and analysing patterns of financial
impersonation across digital plat-
forms.
Table 2: Technological and Algorithmic Countermeasures to Fi-
nancial Crimes.

The above classification matches the technological countermeasures, the ML Countermeasure

Approaches and the Examples of Quantum Algorithms that can be used for each case as it is
disclosed in Table 2.

2 Advantages of Quantum Artificial Intelligence (QAI) Over

Conventional AI Methods
Quantum Artificial Intelligence (QAI) offers several advantages over conventional Artificial Intel-
ligence (AI) methods [4]. Quantum algorithms provide exponential speedups for ML, simulation,
and optimisation tasks, enabling real-time or predictive analysis of significantly larger datasets.
Additionally, QAI can solve problems that are intractable for classical computers, such as integer
factorisation and certain machine learning tasks, due to the exponential resource requirements of
these problems. Moreover, the massive parallelism of quantum systems allows QAI models to poten-
tially recognise complex patterns across vast amounts of data, surpassing the capabilities of classical
AI. Quantum techniques like amplitude amplification and quantum dimensionality reduction excel
at automatically discovering meaningful features without human input. Quantum representations
can encode highly non-linear and intricate relationships between variables, often present in real-
world data. Furthermore, entanglements of quantum systems have the potential to provide new
intuitive insights about data beyond what classical models offer. Certain QAI algorithms, like vari-
ational circuits, can maintain advantages even in near-term device noise.
Additionally, combining quantum and classical techniques into hybrid systems may yield capabil-
ities greater than either paradigm alone. QAI offers performance, representational, and insightful
benefits compared to conventional AI methods. Its speed, scalability, pattern recognition abilities,
feature discovery techniques, representation of non-linear relationships, potential for new insights,
noise tolerance, and hybrid approaches all contribute to its promise in advancing AI capabilities.
In addition to the above, several metrics can be used for comparing QAI to traditional ML, such
as Accuracy, Training time, Inference time, Data size, Scalability, Generalization, Flexibility, Ex-
plainability, Energy efficiency, and Cost of training/deployment [5, 6].
Firstly, in terms of accuracy, QAI can potentially achieve higher accuracy with larger and more
complex datasets, thanks to quantum advantages. Secondly, QAI models could significantly reduce
training time by leveraging quantum acceleration or amplification. Additionally, QAI may enable
near-instant inference on large datasets, thereby improving inference time. Moreover, QAI has
the potential to effectively process massive datasets that surpass the limits of classical computing,
addressing big data problems. Furthermore, quantum parallelism could exponentially enhance the
scalability of QAI models, allowing for the incorporation of more training data and parameters.
QAI models may also exhibit better generalisation by enhancing pattern recognition capabilities.
With quantum computational flexibility, QAI models can be easily adapted or retrained for new

7
tasks, enabling highly versatile lifelong learning models. In terms of explainability, certain QAI
approaches, such as variational circuits, may offer more interpretability compared to deep learning
models. Additionally, QAI’s inherent parallelism can lead to significantly lower power consumption,
making it more energy-efficient for specific tasks. Finally, while there may be initial costs associ-
ated with hardware, expertise, and infrastructure, the cost of training and deploying QAI models
is expected to decrease over time as technology progresses, making it a more economical choice.

3 Unleashing the Power of Quantum Computing in Finan-

cial Crime
Quantum computing offers a range of powerful applications in financial crime detection and pre-
vention. QML models, such as Quantum Neural Network (QNN) and Quantum Support Vector
Machine (QSVM), have the potential to analyse massive financial transaction datasets much faster
than classical AI methods [7]. They can identify complex money laundering or fraud schemes that
might go unnoticed by conventional approaches. Quantum Natural Language Processing (QNLP)
can swiftly sift through various texts, including reports and calls, to detect signs of economic crimes
by better understanding semantics and contextual concepts [8, 9].
Quantum Reinforcement Learning (QRL) agents can interact with synthetic environments that sim-
ulate financial systems at scale [10, 11]. It enables them to identify new vulnerabilities exploited by
criminals, providing valuable insights for strengthening security measures. Quantum recommenda-
tion systems can learn customer behaviours and promptly detect anomalous activities [12], allowing
for immediate flagging and investigation before significant losses occur. Furthermore, quantum di-
mensionality reduction techniques extract meaningful features from high-dimensional time series
and other real-world financial data. It enhances the ability to accurately detect anomalies and
suspicious patterns [13]. Quantum clustering algorithms also play a significant role by grouping
related criminal activities or entities that transcend borders or employ dissimilar techniques [14],
providing a more comprehensive understanding of interconnected criminal networks.
Quantum simulations of markets and investment strategies offer insights into systemic risks and
allow for testing interventions at a speed and scale that classical computers cannot match [15].
Finally, combining the advantages of both paradigms, hybrid quantum-classical models can po-
tentially transform how financial intelligence agencies and banks leverage data to prevent financial
harm. By harnessing the power of quantum computing alongside classical methods [16], these hybrid
models can revolutionise the approach to combating financial crimes. In summary, the applications
of quantum computing in financial crime detection and prevention are vast and promising. From
analysing large datasets to understanding complex semantics and detecting anomalies in real-time,
quantum computing is poised to reshape the context of safeguarding financial systems and prevent-
ing illicit activities.
As mentioned before, QML algorithms, such as QSVM, can revolutionise the analysis of large
transaction datasets by detecting patterns that indicate money laundering, fraud, and other il-
licit activities [17, 18]. The quantum version of these algorithms can perform the classification
task exponentially faster. In addition, quantum annealing and Variational Quantum Eigensolver
(VQE) algorithms offer the opportunity to optimise risk models for extensive portfolios, enabling
the detection of abnormal financial activities associated with economic crimes like insider trad-
ing. Quantum Principal Component Analysis (QPCA) and other dimension reduction techniques
can extract meaningful features from multidimensional financial time series, aiding in identifying

8
anomalies and suspicious transactions related to theft and embezzlement. Furthermore, quantum
order finding algorithms, such as Shor’s algorithm [19], can analyze periodic patterns in finan-
cial market indicators and transactions, uncovering cyclical behaviour exploited by perpetrators
of economic crimes such as pump-and-dump schemes. By leveraging quantum graph algorithms,
particularly quantum walk-based algorithms applied to transaction networks, hidden connections
between entities involved in economic crimes can be detected faster than with classical methods
[20]. Quantum simulations of financial processes and models have the potential to accelerate stress
testing of controls and risk analysis, facilitating the identification of new vulnerabilities exploited
by economic criminals [21]. Moreover, quantum sampling abilities can generate realistic synthetic
economic crime datasets, which can be used to train machine learning models for detection and
prevention. Finally, combining hybrid quantum-classical algorithms with investigative techniques
and domain expertise holds the promise of revolutionising economic crime fighting. By harnessing
the power of quantum computing alongside classical approaches, we can potentially enhance our
ability to combat and prevent economic crimes more effectively.
Financial crimes such as money laundering, terrorist financing, fraud, and tax evasion have signif-
icant financial implications for society, costing billions of dollars each year. These illicit activities
erode trust in the financial system and facilitate more severe criminal behaviours. In the past, finan-
cial institutions primarily relied on human analysts to monitor transactions and report suspicious
activities. However, modern financial systems’ scale, complexity, and speed have surpassed human
capabilities. Enormous volumes of data are generated daily, holding the potential to reveal crimi-
nal patterns. Advanced analytics and AI offer promising solutions to this challenge. By processing
vast amounts of unstructured data, including transactions, documents, and communications, AI
technologies can augment human analysts and identify intricate patterns and anomalies that might
go unnoticed. Equipped with the right tools and access to aggregated transaction data, regulators
and law enforcement agencies can more effectively detect criminal networks, track illicit financial
flows, and prevent unlawful funds from infiltrating the financial system.
From a business perspective, financial firms aim to fortify their platforms against criminal ex-
ploitation, minimise fines and reputational risks, and assure customers that their assets are secure.
Advanced monitoring and AI-based systems safeguard the interests of all stakeholders involved.
Moreover, the application of these new technologies holds great promise in enabling authorities to
swiftly and efficiently identify threats to public safety, such as terrorism, human trafficking, cor-
ruption, and other forms of organised criminal schemes that undermine society as a whole.
QML models, such as QNN and QSVM, offer the ability to analyse massive financial transaction
datasets at a much faster pace than classical AI methods. These models excel in identifying com-
plex money laundering or fraud schemes that may elude traditional approaches. Quantum natural
language processing enables quick sifting through texts like reports and calls, leveraging a bet-
ter understanding of semantics and contextual concepts to detect signs of economic crimes more
effectively [22, 23]. Furthermore, quantum reinforcement learning agents can interact with syn-
thetic environments that simulate financial systems on a large scale. It allows them to identify new
vulnerabilities exploited by criminals, providing valuable insights for enhancing security measures.
Quantum recommendation systems learn customer behaviours and promptly detect anomalous ac-
tivities, enabling immediate flagging for investigation, thus mitigating potential significant losses.
Quantum dimensionality reduction techniques extract meaningful features from high-dimensional
time series and other real-world financial data, improving anomaly detection capabilities. Quantum
clustering algorithms comprehensively group related criminal activities or entities that transcend
borders or employ dissimilar techniques, enhancing the understanding of interconnected criminal

9
networks. Moreover, quantum simulations of markets and investment strategies contribute to a
deeper comprehension of systemic risks and enable testing interventions at a speed and scale not
achievable with classical computers. Finally, hybrid quantum-classical models, harnessing both
paradigms’ advantages, can fundamentally transform how financial intelligence agencies and banks
leverage data to prevent financial harm. In summary, quantum computing applications in financial
crime detection and prevention encompass many capabilities. From analysing massive datasets
and understanding complex semantics to identifying vulnerabilities and mitigating risks, quantum
computing has the potential to revolutionise the way we safeguard financial systems and combat
illicit activities.
The effective detection of financial crimes relies on the utilisation of advanced analytics, AI, and
extensive financial data to uncover both familiar and novel manipulation techniques employed by
criminal organisations. Transaction monitoring plays a vital role in identifying suspicious patterns,
including atypical fund transfers, deposits/withdrawals, and mingling of personal and business
funds, as well as structuring tactics to evade reporting thresholds. Customer due diligence is es-
sential during onboarding to establish risk profiles, while ongoing monitoring ensures that these
profiles remain consistent with transactional activities. Inconsistencies and suspicious identification
documents serve as red flags. Entity resolution enables linking interconnected accounts, entities,
and individuals that may be intentionally disguised, thereby exposing larger schemes and criminal
networks.
Network analysis facilitates the mapping of connections between transacting parties by exploring
common accounts, addresses, and phone numbers, revealing clusters and intermediaries indicative
of underground banking operations. Fund flow analysis traces the trajectory of funds over time
through intricate networks, shedding light on the funds’ ultimate source, destination, and pur-
pose, thereby unravelling layers of transactions designed to obfuscate illicit activities. Behavioural
analysis involves constructing models of normal customer behaviours and establishing transaction
benchmarks, enabling the identification of significant deviations such as changes in transaction loca-
tions or volumes that deviate from established patterns and triggering alerts. Predictive modelling
leverages financial and contextual data to develop risk scores and prioritise alerts, empowering
investigators to focus on higher-risk potential leads. Finally, data aggregation from internal and
external sources provides valuable context concerning customers and geopolitical risks, facilitating
comprehensive risk assessments. By harnessing the power of advanced analytics, AI, and extensive
data, financial institutions and law enforcement agencies enhance their ability to detect and combat
financial crimes, both well-known and innovative, perpetrated by criminal organisations.
QML algorithms, such as QSVM and quantum Boltzmann machines, offer powerful tools for
analysing patterns in large databases to identify anomalies indicative of crimes like fraud, em-
bezzlement, and insider trading [24, 25]. These algorithms can process vast amounts of data and
detect subtle deviations more efficiently than classical methods by leveraging the computational
advantages of quantum computing. Additionally, quantum graph algorithms, specifically quantum
walks applied to networks of entities and transactions, can potentially uncover hidden connections
between perpetrators of complex financial crimes at an accelerated pace [26]. It enables the detec-
tion of intricate relationships and networks that may not be readily apparent using conventional
approaches [7]. Quantum clustering algorithms [27] provide an automated means of grouping re-
lated financial crime activities involving different entities and accounts across various locations and
time frames. By identifying common patterns and associations, these algorithms contribute a more
comprehensive understanding of criminal activities.
Quantum Order Finding (QOF) and period detection algorithms, such as Shor’s algorithm, analyse

10
periodic transaction behaviours, allowing for the identification of activities such as Ponzi schemes
and pump-and-dump operations. It enables timely detection and intervention in fraudulent prac-
tices. Quantum dimension reduction techniques, such as PCA and Singular Value Decomposition
(SVD), extract meaningful financial features that can be used to detect outliers and flag suspicious
activities for further investigation. These techniques improve anomaly detection capabilities and
enhance the accuracy of crime identification. Quantum state and process tomography enable rigor-
ous testing of controls and the identification of weaknesses that financial cybercriminals may exploit
[28]. By conducting simulations and analysing the behaviour of financial systems, these techniques
provide valuable insights for strengthening security measures.
Hybrid quantum-classical optimisation approaches can accelerate risk modelling of large portfo-
lios and databases, enabling the detection of abnormal flows that may indicate financial crimes.
By leveraging both quantum and classical computing capabilities, these models improve the effi-
ciency and accuracy of risk assessment. Quantum generative models offer the ability to expand
limited financial crime datasets synthetically. By generating realistic data, these models facilitate
the robust training of classical and QML algorithms, improving their performance in identifying
and preventing financial crimes. Furthermore, quantum analogue and digital simulation techniques
aid in modelling the dynamics of financial systems at scale. It provides a deeper understanding of
macro patterns exploited by organised crime syndicates, aiding in developing more effective preven-
tion and intervention strategies. In summary, quantum algorithms’ speed, scale, and parallelism
significantly enhance the capabilities of detecting diverse financial crimes. By leveraging quantum
machine learning, graph algorithms, dimension reduction techniques, order finding, state tomog-
raphy, hybrid optimisation, generative models, and simulation methods, financial institutions can
strengthen their defences against criminal activities and protect the integrity of financial systems.

4 Quantum Computing for Malicious Finance Detection: Ben-

efits and Evaluation Metrics
Quantum computing offers significant benefits for the detection of malicious finance activities.
First, quantum algorithms provide exponential speedups, enabling near real-time analysis of huge
datasets impractical for classical computers [29]. Additionally, quantum computing efficiently han-
dles datasets and systems of a scale that exceeds the capabilities of classical approaches [30]. Fur-
thermore, quantum approaches can potentially expose complex correlations and patterns hidden in
classical methods, enhancing detection accuracy [31]. The speed of quantum computing also en-
ables interactive exploration of edge cases and facilitates what-if analysis, offering insights that are
not feasible to obtain classically. Moreover, hybrid quantum-classical tools may provide valuable
insights into the origins and mechanics of detected activities, contributing to their explainability.
Various metrics can be considered to evaluate the effectiveness of quantum-based malicious finance
detection properly. One important metric is the time to detection, which quantifies the reduction
in the time taken to identify malicious activities using quantum analysis compared to classical
methods. Another metric is the detection rate, which measures the percentage of simulated or his-
torical attempts that quantum models successfully flag compared to classical approaches. Striking
a balance between detection coverage and minimising unnecessary alerts is crucial, and this can be
assessed through the false positive rate metric. Additionally, evaluating the impact of additional
or refined data fields and features on detection performance, known as feature sensitivity, helps
optimise the detection process. The robustness of the models against adversarial manipulation of

11
data or strategies aimed at avoiding detection is another important metric to consider.
Scalability metrics focus on determining the maximum volume of data, entities, and interconnections
that can be effectively analysed within a given timeframe using quantum computing. Furthermore,
the fidelity and interpretability of explanations provided by quantum models can be quantitatively
assessed to aid in investigating detected malicious activities. Finally, the capability of quantum
computing to identify previously unknown activities or schemes beyond what classical methods can
achieve, known as novelty, is an essential metric to consider. Proper evaluation against these bene-
fits and metrics is crucial to validate the value proposition of quantum computing in the context of
detecting and combating malicious finance activities. It enables a comprehensive assessment of the
effectiveness, efficiency, and practical applicability of quantum approaches, ultimately supporting
informed decision-making regarding the adoption of quantum solutions in this domain.

5 Enhancing Money Laundering Detection with Quantum

Algorithms
The application of financial analytics, network analysis, and machine learning plays a crucial role
in detecting patterns, relationships, and risk factors that may go unnoticed by humans but could
potentially indicate money laundering or other forms of financial crime. Various techniques are
employed for this purpose, including transaction monitoring, customer due diligence, funds flow
analysis, behavioural analysis, network analysis, transaction benchmarking, transaction structur-
ing detection, predictive modelling, entity resolution, and data aggregation.
Transaction monitoring involves scrutinising wire transfers, deposits, withdrawals, and other activ-
ities for suspicious patterns, including cross-border flows. Customer due diligence entails screening
customers against watchlists and sanction lists and monitoring transactions to ensure consistency
with customer risk profiles. Fund flow analysis focuses on tracing the movement of funds across
interconnected accounts, corporations, and individuals to unveil money laundering schemes. Be-
havioural analysis utilises machine learning to establish profiles of typical account usage and identify
anomalous transactions that deviate from established behaviours. Network analysis aims to map
connections between transacting parties, identifying clusters, common intermediaries, and under-
ground banking systems.
Transaction benchmarking involves comparing transaction amounts, frequencies, geographies, and
counterparties against peer groups to flag outliers. Transaction structuring detection aims to iden-
tify the intentional fragmentation of transactions to evade reporting requirements. Predictive mod-
elling involves developing models that assess transaction risks and prioritise alerts for investigators.
Entity resolution focuses on linking related transactions, accounts, and identities that may be in-
tentionally disguised. Finally, data aggregation combines internal and external data sources to gain
a more comprehensive view of risks. By leveraging these techniques and technologies, financial
institutions and law enforcement agencies can enhance their ability to combat financial crime effec-
tively.
As mentioned before, QML algorithms, such as QSVMs, can analyse large transaction datasets
and swiftly identify suspicious patterns indicative of money laundering. These quantum algorithms
outperform their classical counterparts, enabling faster and more accurate classification tasks. Ad-
ditionally, quantum graph algorithms, specifically quantum walks applied to transaction networks,
excel in detecting hidden connections between entities and uncovering complex money laundering
rings or schemes at an accelerated pace compared to classical algorithms [32, 22].

12
Quantum clustering algorithms are invaluable in automatically grouping related money laundering
activities based on transaction attributes, revealing connections that may not be apparent through
traditional means. Quantum dimensionality reduction techniques, such as QPCA, extract mean-
ingful features from high-dimensional financial data. It facilitates the detection of anomalies and
the flagging of suspicious transactions for further investigation. Furthermore, the quantum order
finding algorithm can analyse periodic or cyclical patterns in financial activities, allowing for the
identification of behaviour typical of specific money laundering techniques like round-tripping or
smurfing. Quantum state tomography, applied to the output of quantum simulations of financial
networks or processes, enables stress testing of controls and the identification of vulnerabilities ex-
ploited by money launderers.
Hybrid quantum-classical optimisation algorithms can potentially accelerate risk modelling of large
and complex portfolios, enabling the detection of abnormal flows that may indicate money launder-
ing activities. Additionally, quantum sampling techniques can aid in generating realistic synthetic
money laundering datasets, augmenting the training of machine learning models and improving
their effectiveness. Indeed, the speed and scalability of quantum algorithms significantly enhance
money laundering detection capabilities. By leveraging quantum machine learning, graph algo-
rithms, dimensionality reduction techniques, order finding algorithms, state tomography, hybrid
optimisation algorithms, and sampling methods, financial institutions can more effectively combat
money laundering and safeguard their systems against illicit activities.

6 Enhancing Cryptocurrency Crime Detection with Quan-

tum Computing
The detection and prevention of financial crimes employ advanced algorithms and leveraging ex-
panding financial crime databases to triangulate structural and behavioural analyses. Screening
transactions against watchlists of known criminals, sanctioned entities, and politically exposed per-
sons is a significant measure to identify potentially corrupted interactions. Analysing transaction
metadata, including payee names, memo fields, and account descriptors, provides valuable context
for categorising transactions and uncovering illicit schemes.
Geotagging techniques play a crucial role in identifying abnormal transaction locations that may
indicate fraudulent activity while monitoring cross-border flows, which assists in detecting potential
movements of illicit funds. Structural analysis delves into examining relationships, communication
patterns, and fund movements within networks to expose hierarchies, connections, and the division
of labour within criminal organisations.
Link and cluster analysis identifies associations between accounts and entities over time, drawing
upon attributes such as overlapping addresses, phone numbers, and IP addresses, exposing hidden
relationships. Machine learning-based anomaly and outlier detection utilise historical databases
of past criminal and non-criminal transactions and patterns to identify statistical deviations that
necessitate further investigation automatically. A enhances and expedites these techniques by con-
tinuously learning from investigators, adapting models, and recommending new monitoring patterns
as criminal methods evolve.Effective coordination and information sharing between the public and
private sectors are also essential in combating financial criminal networks. This collaborative ap-
proach ensures the efficient exchange of intelligence and maximises the collective effort to combat
financial crimes.
As mentioned above, QML algorithms, particularly QSVM, offer powerful tools for analysing pat-

13
terns in graphs of cryptocurrency transactions and addresses [18]. These algorithms can swiftly flag
suspicious entities and activities indicative of theft, scamming, and money laundering. By lever-
aging the computational advantages of quantum computing, these algorithms outperform classical
methods, enabling faster detection of illicit activities in the crypto space [33].
Quantum period finding algorithms contribute to the identification of repetitive patterns in cryp-
tocurrency transfers, which often signify pump-and-dump schemes or round-tripping activities as-
sociated with market manipulation crimes. This capability assists in the timely detection and
prevention of fraudulent practices [34]. Quantum clustering algorithms excel in grouping crypto
transactions and entities involved in complex crimes that span multiple accounts and currencies,
overcoming the limitations of classical clustering techniques. This comprehensive approach en-
hances understanding of interconnected criminal networks within the cryptocurrency ecosystem.
Quantum signature detection techniques analyse various attributes of crypto transactions, such as
coin amounts and velocities, to identify anomalies and outlier transactions associated with theft
and fraud crimes [25]. By leveraging quantum computing’s computational power, these algorithms
can efficiently detect suspicious activities that may go unnoticed by traditional methods.
Quantum dimension reduction methods, such as PCA and SVD, extract meaningful financial fea-
tures from high-volume and complex cryptocurrency transaction data. It enables the identification
of outliers and flags transactions for further investigation, enhancing anomaly detection capabili-
ties. Cryptocurrency network analysis, utilising quantum walk algorithms applied to networks of
entities and transactions, uncover hidden connections between players involved in syndicated crypto
crimes. By revealing intricate relationships, these algorithms provide valuable insights for combat-
ing organised criminal activities within the crypto space. Quantum simulation techniques allow for
modelling the dynamics of large cryptocurrency markets and systems at a scale that classical com-
puters cannot match. It provides a deeper understanding of macro patterns exploited by organised
criminal networks, aiding in developing effective prevention and intervention strategies.
Hybrid quantum-classical tools integrate the results obtained from quantum analytics with tradi-
tional forensics, enabling improved detection of cryptocurrency crimes. By combining the strengths
of both paradigms, these tools enhance the accuracy and efficiency of crime detection and investi-
gation in the crypto domain. In summary, the parallelism and scalability of quantum computing
significantly enhance the capabilities of detecting and combating cryptocurrency crimes. Through
the utilisation of quantum machine learning, period finding, clustering, signature detection, dimen-
sion reduction, network analysis, simulation, and hybrid quantum-classical tools, the cryptocur-
rency ecosystem can be safeguarded against illicit activities, protecting investors and maintaining
the integrity of the market.

7 Enhancing Detection of Market Manipulation with Quan-

tum Computing
Unravelling manipulative schemes that are often intentionally designed to evade detection requires
the implementation of sophisticated algorithms, pattern-matching techniques, and entity resolution.
Manipulative practices such as identifying wash trades must be addressed, which involve fake trans-
actions to create misleading activity or liquidity in a stock without changing beneficial ownership.
Monitoring for spoofing is essential, which entails entering large orders and cancelling them to arti-
ficially influence price movements before executing trades. Additionally, detecting pump-and-dump
schemes is crucial, whereby coordinated buying artificially inflates the price of a stock, followed by

14
selling to dispose of shares on unsuspecting retail investors.
Analysing the concentration of trades is necessary to identify instances where one entity dominates
the trading volume in a stock, particularly around news announcements or filing periods. Compar-
ing trade volumes to the public float is important, as abnormally high volumes could indicate the
trading of fake shares. Benchmarking price moves against news events is also crucial, as drastic
price changes without corresponding news raise concerns. The behavioural analysis of traders plays
a significant role in identifying coordinated patterns among entities and using the same brokerages
or accounts.
Assessing order-to-trade ratios is essential, as high proportions of cancelled orders compared to
completed trades could signal spoofing. Network analysis of trading relationships helps identify
dense clusters of entities repeatedly engaging in trades with each other. Machine learning tech-
niques can be employed, using past cases to train models to detect newer manipulation techniques
autonomously. Ongoing monitoring and the adaptability of models are vital to keep pace with the
evolving tactics employed by manipulators. By leveraging these sophisticated tools and techniques,
it becomes possible to detect and mitigate manipulative activities in financial markets effectively.
We have already mentioned that QML algorithms, such as QSVM, offer powerful tools for analysing
patterns in large datasets of stock trades, orders, and prices [35, 24]. By leveraging these algorithms,
market manipulation strategies like pump-and-dump and spoofing can be automatically identified
faster. The computational advantages of quantum computing enable efficient processing of vast
amounts of data, improving the detection of unfair practices in the stock market. Quantum graph
analysis techniques, particularly quantum walks applied to trading entities and transaction net-
works, can uncover complex insider trading rings and hidden coordination between actors. By
examining the relationships and interactions within the market, these algorithms provide valuable
insights into illicit activities that may not be easily detectable using traditional methods [36].
Quantum dimension reduction methods, such as Quantum Principal component analysis (QPCA)
and Qunatum SVD (QSVD), extract meaningful features from high-dimensional market data. It
enables the identification of anomalous trades, prices, or volatility patterns that may indicate mar-
ket manipulation. Through quantum computing’s computational power, these techniques enhance
the accuracy of identifying suspicious activities.
Quantum signal processing techniques focus on detecting engineered price or volume signals that
could indicate front-running or insider trading activities. By analysing stock market signals in the
quantum domain, these algorithms can identify patterns that may not be apparent using classical
methods, contributing to detecting unfair practices in real-time. Quantum order-finding algorithms
analyse repetitive or cyclical trading periodicities, which may wash trading or other strategies
designed to influence markets artificially. By leveraging quantum computing’s capabilities, these
algorithms provide insights into trading patterns that can assist in exposing market manipulation
techniques.
Quantum portfolio optimisation techniques allow stress tests on large and complex portfolio hold-
ings using quantum solvers. It enables verifying compliance with regulations, particularly regarding
conflicts of interest, enhancing transparency and accountability in the financial industry. Quantum
simulation techniques enable the modelling of stock exchange dynamics and the co-movement of
related firms’ prices at scale. By gaining insights into potential market manipulation techniques,
these simulations contribute to a deeper understanding of unfair practices in financial markets.

15
8 Quantum Circuit Born Machines in Finance: Enhancing
Generative Modeling and Quantitative Applications
A Quantum Circuit Born Machine (QCBM) is a QML model based on Born machines and offers
several key advantages [37]. Born machines are variational hybrid quantum-classical models that
leverage Born’s rule from quantum mechanics to evaluate classical data distributions. In the case of
QCBMs, the Born rule distribution is implemented using a parameterised quantum circuit, enabling
the model to harness the power of quantum hardware [38].
The quantum circuit encodes a quantum state representing the data distribution, and measure-
ments in this state provide samples from the distribution. The circuit parameters are optimised
variationally using a classical optimiser to minimise the difference between the generated and target
distributions. This approach offers a quantum-enhanced generative modelling capability that can
efficiently learn certain distributions compared to classical methods. QCBMs have demonstrated
their potential in various tasks, including density estimation, molecular simulations, graph neural
networks, and generative adversarial networks. One of the advantages of QCBMs is that they have
relatively weaker resource requirements than other QML models, making them implementable on
near-term quantum devices. However, it is important to note that their practical implementation
is still constrained by the size of circuits that can be realised.
In finance, QCBMs hold promise for various applications [39]. For instance, in option pricing and
hedging, QCBMs can model complex underlying asset distributions more accurately than classical
methods, enabling the pricing of exotic options and effective risk management of large option port-
folios.
QCBMs can also aid in portfolio optimisation by capturing the intricate covariances between assets
and providing recommendations for optimal asset allocations. Risk modelling becomes more refined
with QCBMs as they can simulate extreme but plausible scenarios from the learned distribution,
assisting in estimating portfolio Value at Risk and Expected Shortfall. Furthermore, QCBMs can
be utilised for anomaly detection in finance, helping to identify outliers and anomalies in time se-
ries data, such as stock prices, by learning normal distributions and flagging deviations. In fraud
detection, QCBMs can model complex patterns in large financial transaction graphs, enabling the
identification of traces of money laundering, spoofing, and pump-and-dump schemes. Macroeco-
nomic forecasting can benefit from QCBMs as they can capture complex non-linear relationships
in economic indicators, facilitating predictions of recessionary environments and commodity prices.
Stress testing of banks’ resilience to market crashes and liquidity crises can be enhanced through
QCBMs by simulating distributions of risk factor shocks.
In trading strategies, QCBMs can aid in developing and evaluating strategies like market making
by modelling order books and limiting order placements. Additionally, QCBMs can be used for
synthetic data generation, producing realistic training datasets by sampling from QCBM distri-
butions to augment classical models. The unique modelling abilities offered by QCBMs have the
potential to improve various quantitative finance applications significantly. By leveraging the power
of quantum computing, QCBMs can provide more accurate and efficient solutions for pricing, risk
management, anomaly detection, fraud detection, forecasting, stress testing, trading strategies, and
synthetic data generation in the financial industry.

16
9 Advancing Risk Management in Finance: Quantum Com-
puting for Enhanced Analysis
Quantum computing holds immense potential for revolutionising risk management in the financial
industry. By leveraging the speed, scale, and parallelism inherent in quantum systems, various
applications can be developed to enhance risk analysis and decision-making. Here are some key
areas quantum computing can significantly impact [40, 41].
Quantum portfolio optimisation allows using quantum optimisation techniques, such as VQE [42],
to efficiently determine optimal asset allocations for large and complex portfolios while consider-
ing risk constraints. This approach has the potential to outperform classical methods and deliver
superior portfolio performance. Quantum scenario analysis introduces quantum simulations to rig-
orously assess risks under extreme market conditions that are challenging to model using classical
approaches. By leveraging the computational power of quantum systems, financial institutions can
gain deeper insights into the potential impact of rare and catastrophic events on their portfolios.
Quantum Value-at-Risk (VaR) estimation involves leveraging QML algorithms, like QSVMs, to es-
timate VaR and Expected Shortfall (ES) risk measures using vast historical market data. Quantum
algorithms can process and analyse this data more efficiently, enabling more accurate and timely
risk assessments. Quantum risk factor modelling employs QPCA techniques to extract risk factors
from high-dimensional market time series data. Financial institutions can improve risk modelling
and management strategies by identifying the underlying factors driving portfolio risk.
Quantum correlation analysis offers the ability to analyse the joint interactions of multiple risk
factors exponentially faster through quantum algorithms. This enhanced capability enables a more
comprehensive and accurate portfolio risk modelling, particularly in complex and interconnected
markets. Quantum counterparty risk analysis involves applying quantum graph algorithms to map
connections between financial entities, identifying potential systemic risks and vulnerabilities within
the financial system.
By leveraging quantum computing’s ability to handle and analyse large-scale networks, financial
institutions and regulators can gain deeper insights into counterparty risk and systemic stabil-
ity. Quantum stress testing utilises hybrid quantum-classical optimisation approaches to test the
resilience of risk models under diverse assumption shocks at a large scale. By combining the com-
putational power of quantum systems with classical optimisation techniques, financial institutions
can assess the robustness of their risk management frameworks and identify areas for improvement.
Quantum anomaly detection leverages QML techniques to flag outlier market moves within massive
datasets. By harnessing the power of quantum systems, financial institutions can enhance their risk
investigation capabilities and promptly address potential anomalies and irregularities. Overall, the
speed, scale, and parallelism offered by quantum computing have the potential to advance risk
management capabilities for financial institutions and regulators significantly. By embracing quan-
tum technologies and developing innovative applications, the financial industry can enhance risk
analysis, improve decision-making processes, and strengthen overall financial stability.

10 Comparing Quantum Algorithms and Classical Approaches:

Key Findings
In comparing classical and quantum computing approaches for detecting financial crimes, the dis-
tinctions largely revolve around speed, data processing, and pattern detection accuracy. Classical

17
computing, the foundation for fighting financial crime, processes information in binary, limiting its
speed and efficiency as data volumes grow. As can be seen in Figure 1, quantum computing, lever-
aging principles like superposition and entanglement, promises significant improvements in these
areas but remains in the early development stages for practical application.

• Processing Speed
• Data Processing Capabilities
Quantum • Accuracy in Detecting Fraud Patterns
Computing
Advantages

• Technological Maturity
• Utllizing Quantum Advantages (such as
superposition and entanglement)
Quantum Future • Integration Challenges
• Resource Intensity
Opportunities and
Challenges

• Quantum Research
• Quantum-Ready Workforce
• Initiate Pilot Projects
• Quantum-Secure Cryptography
Recommendations • Regulatory Frameworks
• Cross-Sector Collaboration
for using Quantum • Monitor Developments in Quantum Computing

in Finance

Figure 1: Quantum Context: Advantages, Opportunities, and Challenges in Finance.

- Processing Speed: Classical methods are bound by algorithms’ linear or polynomial time com-
plexities, which can slow searching through massive datasets. Using algorithms like Grover’s,
Quantum computing could dramatically reduce search times, offering a quadratic speedup.
Indeed, quantum computers could potentially complete tasks in hours or days that would take
classical computers much longer.
- Data Processing Capabilities: Classical computing has advanced with big data technologies
and ML, enabling the processing of large datasets to identify fraud patterns. However, the
increasing volume and complexity of data present challenges. Using techniques like QPCA,
Quantum computing could theoretically manage high-dimensional data more efficiently, taking
advantage of quantum computing’s ability to operate in high-dimensional spaces.
- Accuracy in Detecting Fraud Patterns: The accuracy of classical AI and ML models in iden-
tifying fraud can be affected by the complexity of financial systems. Quantum computing’s
approach to processing multiple possibilities simultaneously could theoretically improve model
accuracy by uncovering subtle correlations missed by classical methods. However, the practi-
cal benefits of quantum computing in enhancing detection accuracy are yet to be fully realised.

18
 The transition from classical to quantum computing in financial crime detection is expected
to bring significant advancements but is contingent on overcoming technological challenges
and developing new algorithms and security measures. The potential for quantum computing
to transform the field is clear, yet much work remains to harness these capabilities fully.
Based on the comparison between classical and quantum computing in detecting financial crimes,
several findings emerge, leading to recommendations for future research, development, and policy
formulation:
a) Technological Maturity, given that quantum computing is nascent, especially in practical
applications for financial crime detection. While promising, the technology requires further
development to realise its full potential;
b) Quantum Advantage: The theoretical advantages of quantum computing—such as processing
speed and the ability to handle complex, high-dimensional data—are clear. However, prac-
tical demonstrations of these advantages in real-world financial crime detection scenarios are
limited;
c) Integration Challenges: Integrating quantum computing into financial crime detection frame-
works poses significant challenges, including the need for new algorithms, data security mea-
sures, and possibly regulatory changes;
d) Resource Intensity: Quantum computing currently requires substantial resources, including
specialised knowledge and quantum computing hardware, which are not widely accessible.
The above analyses provided some useful recommendations:
a) Investment in Research and Development: Governments, academia, and the private sector
should invest in R&D to advance quantum computing technology, focusing on developing
practical applications for financial crime detection;
b) Develop Quantum-Ready Workforce: Educational institutions and industry players should
work together to develop training programs that prepare a new generation of professionals
skilled in quantum computing and its applications in finance;
c) Pilot Projects: In collaboration with quantum computing firms, financial institutions should
initiate pilot projects to explore the application of quantum computing in detecting financial
crimes. These projects can provide valuable insights into the technology’s practical benefits
and limitations;
d) Quantum-Secure Cryptography: With the advancement of quantum computing, existing en-
cryption methods could become vulnerable. Developing quantum-secure cryptography is es-
sential to protect financial data against potential quantum computing threats;
e) Regulatory Frameworks: Regulatory bodies should begin considering the implications of quan-
tum computing on financial systems. Developing guidelines and standards that address the
use of quantum computing in finance will be crucial to ensuring security, privacy, and fair
use;
f) Cross-Sector Collaboration: Encourage collaboration across financial institutions, technology
companies, regulatory bodies, and academic researchers to share knowledge, resources, and
best practices in applying quantum computing;

19
 g) Monitor Developments in Quantum Computing: Financial institutions should continuously
monitor advancements in quantum computing to stay ahead of potential threats and op-
portunities. It includes keeping abreast of developments in quantum algorithms that could
revolutionise financial crime detection.

11 Conclusions
The development of quantum computing introduces significant advancements in detecting and coun-
tering financial crimes. This technology, leveraging the unique capabilities of quantum mechanics,
offers a new set of tools for analysing data and identifying illicit activities within vast datasets,
doing so at a pace and level of complexity that classical computing can not match. The examples
of quantum algorithms explored in this article, from QSVM for anomaly detection to quantum
clustering for identifying hidden networks involved in money laundering, showcase the potential
applications across various financial crime categories. Quantum computing’s edge lies in its ability
to process information through mechanisms such as superposition and entanglement, providing a
leap in the speed and efficiency of data analysis. This ability is critical in the financial sector, where
detecting fraudulent patterns swiftly can significantly disrupt criminal operations. By integrating
quantum technologies with existing frameworks and strategies for financial crime prevention, there
is a clear path towards enhancing the effectiveness of these measures. The alignment of financial
crime classifications with recognised standards from global entities, such as the FATF and the UN-
ODC, emphasises the need for a unified strategy incorporating quantum computing. This approach
improves existing methods and introduces new capabilities for uncovering and combating financial
crimes. The financial sector faces an opportunity to strengthen defences against criminal activi-
ties. It includes investment in research, development, and collaborative efforts to explore and apply
quantum technologies. The adoption of quantum computing in financial crime detection signals a
move towards more secure, transparent, and resilient financial systems equipped to counter the chal-
lenges posed by sophisticated criminal schemes. This article encourages the financial community to
seriously consider the potential of quantum computing. We can expect a significant enhancement
in the mechanisms for preventing and detecting financial crimes safeguarding the financial system
against emerging threats by adopting this technology.

References
[1] Darı́o Gil. Institute for Business Value (2023) The Quantum Decade: A Playbook for
Achieving Awareness, Readiness, and Advantage, 4th edn. IBM. https://www.ibm.com/
thought-leadership/institute-business-value/en-us/report/quantum-decade, 2023.
[Online; accessed 22-March-2024].
[2] Stefano Mangini, Francesco Tacchino, Dario Gerace, Daniele Bajoni, and Chiara Macchiavello.
Quantum computing models for artificial neural networks. Europhysics Letters, 134(1):10002,
2021.
[3] Sri Amit Ray. Quantum Machine Learning with Quantum Cheshire Cat Generative AI Model:
Quantum Mirage Data. Compassionate AI Lab, 2024.
[4] Vedran Dunjko and Hans J Briegel. Machine learning & artificial intelligence in the quantum
domain: a review of recent progress. Reports on Progress in Physics, 81(7):074001, 2018.

20
 [5] Sukhpal Singh Gill, Minxian Xu, Carlo Ottaviani, Panos Patros, Rami Bahsoon, Arash
Shaghaghi, Muhammed Golec, Vlado Stankovski, Huaming Wu, Ajith Abraham, et al. Ai
for next generation computing: Emerging trends and future directions. Internet of Things,
19:100514, 2022.
[6] Lidong Wang and Cheryl Ann Alexander. Quantum science and quantum technology: Progress
and challenges. Am. J. Electr. Electron. Eng., 8(2):43–50, 2020.
[7] Nouhaila Innan, Muhammad Al-Zafar Khan, and Mohamed Bennai. Financial fraud detection:
a comparative study of quantum machine learning models. International Journal of Quantum
Information, page 2350044, 2023.
[8] Raffaele Guarasci, Giuseppe De Pietro, and Massimo Esposito. Quantum natural language
processing: Challenges and opportunities. Applied Sciences, 12(11):5651, 2022.
[9] Smita Roy Trivedi, Dipali Krishnakumar, and Richa Verma Bajaj. Loan frauds and bad boy
billionaires: A new approach of loan fraud prevention using natural language processing (nlp).
NIBM Working Paper Series, 2021.
[10] Murphy Yuezhen Niu, Sergio Boixo, Vadim N Smelyanskiy, and Hartmut Neven. Universal
quantum control through deep reinforcement learning. npj Quantum Information, 5(1):33,
2019.
[11] Nico Meyer, Christian Ufrecht, Maniraman Periyasamy, Daniel D Scherer, Axel Plinge,
and Christopher Mutschler. A survey on quantum reinforcement learning. arXiv preprint
arXiv:2211.03464, 2022.
[12] Iordanis Kerenidis and Anupam Prakash. Quantum recommendation systems. arXiv preprint
arXiv:1603.08675, 2016.
[13] Iris Cong and Luming Duan. Quantum discriminant analysis for dimensionality reduction and
classification. New Journal of Physics, 18(7):073011, 2016.
[14] Esma Aı̈meur, Gilles Brassard, and Sébastien Gambs. Quantum clustering algorithms. In
Proceedings of the 24th international conference on machine learning, pages 1–8, 2007.
[15] Andrew J Daley, Immanuel Bloch, Christian Kokail, Stuart Flannigan, Natalie Pearson,
Matthias Troyer, and Peter Zoller. Practical quantum advantage in quantum simulation.
Nature, 607(7920):667–676, 2022.
[16] Gennaro De Luca. A survey of nisq era hybrid quantum-classical machine learning research.
Journal of Artificial Intelligence and Technology, 2(1):9–15, 2022.
[17] Juraj Nosál. Crime in the digital age: A new frontier. In The Implications of Emerging
Technologies in the Euro-Atlantic Space: Views from the Younger Generation Leaders Network,
pages 177–193. Springer, 2023.
[18] Rosario Girasa and Gino J Scalabrini. Regulation of Innovative Technologies: Blockchain,
Artificial Intelligence and Quantum Computing. Springer Nature, 2022.
[19] Andrew M Childs and Jason M Eisenberg. Quantum algorithms for subset finding. arXiv
preprint quant-ph/0311038, 2003.

21
[20] Angad Kalra, Faisal Qureshi, and Michael Tisi. Portfolio asset identification using graph
algorithms on a quantum annealer. Available at SSRN 3333537, 2018.
[21] Alexander M Dalzell, Sam McArdle, Mario Berta, Przemyslaw Bienias, Chi-Fang Chen, András
Gilyén, Connor T Hann, Michael J Kastoryano, Emil T Khabiboulline, Aleksander Kubica,
et al. Quantum algorithms: A survey of applications and end-to-end complexities. arXiv
preprint arXiv:2310.03011, 2023.
[22] Michele Grossi, Noelle Ibrahim, Voica Radescu, Robert Loredo, Kirsten Voigt, Constantin
Von Altrock, and Andreas Rudnik. Mixed quantum–classical method for fraud detection with
quantum feature selection. IEEE Transactions on Quantum Engineering, 3:1–12, 2022.
[23] Abiodun Esther Omolara, Aman Jantan, Oludare Isaac Abiodun, Manmeet Mahinderjit Singh,
Mohammed Anbar, and DV Kemi. State-of-the-art in big data application techniques to
financial crime: a survey. International Journal of Computer Science and Network Security,
18(7):6–16, 2018.
[24] Daniel J Egger, Claudio Gambella, Jakub Marecek, Scott McFaddin, Martin Mevissen, Rudy
Raymond, Andrea Simonetto, Stefan Woerner, and Elena Yndurain. Quantum computing for
finance: State-of-the-art and future prospects. IEEE Transactions on Quantum Engineering,
1:1–24, 2020.
[25] Mingchao Guo, Hailing Liu, Yongmei Li, Wenmin Li, Fei Gao, Sujuan Qin, and Qiaoyan Wen.
Quantum algorithms for anomaly detection using amplitude estimation. Physica A: Statistical
Mechanics and its Applications, 604:127936, 2022.

[26] Karuna Kadian, Sunita Garhwal, and Ajay Kumar. Quantum walk and its application domains:
A systematic review. Computer Science Review, 41:100419, 2021.
[27] Somayeh Bakhtiari Ramezani, Alexander Sommers, Harish Kumar Manchukonda, Shahram
Rahimi, and Amin Amirlatifi. Machine learning algorithms in quantum computing: A survey.
In 2020 International joint conference on neural networks (IJCNN), pages 1–8. IEEE, 2020.
[28] Gregory AL White, Felix A Pollock, Lloyd CL Hollenberg, Kavan Modi, and Charles D Hill.
Non-markovian quantum process tomography. PRX Quantum, 3(2):020344, 2022.
[29] Abha Naik, Esra Yeniaras, Gerhard Hellstern, Grishma Prasad, and Sanjay Kumar
Lalta Prasad Vishwakarma. From portfolio optimization to quantum blockchain and secu-
rity: A systematic review of quantum computing in finance. arXiv preprint arXiv:2307.01155,
2023.
[30] Maryann Njorbuenwu, Bobby Swar, and Pavol Zavarsky. A survey on the impacts of quantum
computers on information security. In 2019 2nd International conference on data intelligence
and security (ICDIS), pages 212–218. IEEE, 2019.

[31] Bilgehan Arslan, Mehtap Ulker, Sedat Akleylek, and Seref Sagiroglu. A study on the use
of quantum computers, risk assessment and security problems. In 2018 6th International
Symposium on Digital Forensic and Security (ISDFS), pages 1–6. IEEE, 2018.

22
[32] Soumya Ranjan Mishra, Hitesh Mohapatra, et al. Enhancing money laundering detection
through machine learning: A comparative study of algorithms and feature selection techniques.
In AI and Blockchain Applications in Industrial Robotics, pages 300–321. IGI Global, 2024.
[33] Apoorva Ganapathy. Quantum computing in high frequency trading and fraud detection.
Engineering International, 9(2):61–72, 2021.

[34] Jorão Gomes, Sajjad Khan, and Davor Svetinovic. Fortifying the blockchain: A systematic
review and classification of post-quantum consensus solutions for enhanced security and re-
silience. IEEE Access, 2023.
[35] Anshul Saxena, Javier Mancilla, Iraitz Montalban, and Christophe Pere. Financial Model-
ing Using Quantum Computing: Design and manage quantum machine learning solutions for
financial analysis and decision making. Packt Publishing Ltd, 2023.
[36] Mritunjay Shall Peelam, Anjaney Asreet Rout, and Vinay Chamola. Quantum computing
applications for internet of things. IET Quantum Communication, 2023.
[37] Javier Alcazar, Vicente Leyton-Ortega, and Alejandro Perdomo-Ortiz. Classical versus quan-
tum models in machine learning: insights from a finance application. Machine Learning:
Science and Technology, 1(3):035003, 2020.
[38] Santanu Ganguly. Implementing quantum generative adversarial network (qgan) and qcbm in
finance. arXiv preprint arXiv:2308.08448, 2023.

[39] Brian Coyle, Maxwell Henderson, Justin Chan Jin Le, Niraj Kumar, Marco Paini, and Elham
Kashefi. Quantum versus classical generative modelling in finance. Quantum Science and
Technology, 6(2):024013, 2021.
[40] Meng-Leong How and Sin-Mei Cheah. Business renaissance: Opportunities and challenges at
the dawn of the quantum computing era. Businesses, 3(4):585–605, 2023.

[41] Stefan Woerner and Daniel J Egger. Quantum risk analysis. npj Quantum Information, 5(1):15,
2019.
[42] Jules Tilly, Hongxiang Chen, Shuxiang Cao, Dario Picozzi, Kanav Setia, Ying Li, Edward
Grant, Leonard Wossnig, Ivan Rungger, George H Booth, et al. The variational quantum
eigensolver: a review of methods and best practices. Physics Reports, 986:1–128, 2022.

23