Table of Contents

Preface...................................................................................................................................................
I. Introduction.....................................................................................................................................
A.Definition of Artificial Intelligence (AI).......................................................................................
B. Overview of Financial Market......................................................................................................
C. Trends of AI applications in the development of Banking Industry........................................
II. Applications of AI in Finance Markets.......................................................................................
A. ANN................................................................................................................................................
B. Expert system.................................................................................................................................
C. Hybrid Intelligence in Finance.....................................................................................................
III. Case study - Vietcombank...........................................................................................................
1. Overview..........................................................................................................................................
2. Application of AI in Vietcombank..............................................................................................
3. Human Oversight.........................................................................................................................
4. Results............................................................................................................................................
IV. Challenges & risks of AI in Finance.........................................................................................
V. Future trend and development of artificial technology in the financial market...................
Conclusion.............................................................................................................................................
References.............................................................................................................................................
 Preface
In an age where technology rapidly redefines boundaries and possibilities, Artificial
Intelligence (AI) has emerged as a crucial force in reshaping the financial markets. The
combination of AI and finance is not merely a trend but a profound evolution that is
gradually changing how financial activities are conducted, analyzed, and understood.
This thesis delves deeply into the role AI plays within financial markets, concentrating on:
1. AI Technologies and Their Applications in Financial Markets:
The first section offers a comprehensive look at the various AI technologies and their
roles in the financial sector, with a focus on three key areas:
 Artificial Neural Networks (ANNs): This section explores exhow ANNs are used
in financial applications such as stock market prediction, portfolio optimization, and risk
management. It also includes case studies highlighting real-world implementations of
ANNs.
 Expert Systems: The aim of this part is to explore the use of expert systems in
automating complex financial tasks and enhancing decision-making processes, focusing on
their ability to improve the analytical depth of financial operations.
 Hybrid Intelligence: This section examines the integration of different AI
methodologies to create hybrid systems that leverage the strengths of various AI
approaches, leading to more sophisticated financial analysis and decision -making.
2. Case Study Analysis:
An in-depth case study from one of the highlighted applications will be selected and
analyzed to provide a practical illustration of AI’s impact on financial markets.
3. Challenges and Risks:
While AI offers numerous advantages, there are inherent challenges and risks involved
in its application to finance. This section addresses those concerns, offering a balanced
view of potential pitfalls and strategies to mitigate the associated risks.
4. Future Trends and Developments:
This segment explores the emerging trends in AI technology and their prospective
impact on the financial markets. It aims to provide insight into how AI may continue to
evolve and further influence the sector in the years to come.
Overall, the thesis seeks to thoroughly explore AI technologies—spanning from
machine learning and algorithmic trading to advanced data analytics and natural language
processing—and their transformative influence on financial markets. By analyzing these
technologies, their practical applications, and the challenges they bring, this research hopes
to offer valuable insights into the ongoing technological evolution within the finance
industry. Furthermore, it aims to provide a forward-looking perspective on the future role
AI is likely to play in reshaping the financial landscape.
I. Introduction

A. Definition of Artificial Intelligence (AI)

The term AI was first introduced by John McCarthy in 1956 at the Dartmouth Conference, the
first conference on AI. In 2007, McCarthy defined AI as "the science and engineering of creating
intelligent machines, especially computer programs, which tasks that involve using computers to
understand human intelligence, but not restricted to biologically observable methods" (McCarthy,
2007). The Financial Stability Board (2017) defines AI as "the theory and development of computer
systems able to perform tasks typically requiring human intelligence."
At its core, AI involves the automation of learning and discovery through data. Unlike
traditional robotic automation, which primarily focuses on automating manual, repetitive tasks, AI
handles complex, high-volume tasks requiring cognitive processes. AI systems use sophisticated
algorithms to analyze data, learn from it, and make decisions with minimal human intervention.
Core Components of AI
1. Data: AI systems rely on large volumes of data, which act as training material for the
underlying algorithms. This data enables the AI to recognize patterns, make predictions, and
perform tasks.
2. Algorithms: At the heart of AI are algorithms—complex mathematical formulas that allow
AI systems to process data and learn from it. These algorithms underpin the system’s ability
to adapt and improve over time.
3. Learning Mechanisms: AI systems incorporate learning mechanisms that allow them to
use the knowledge gained from data to enhance their performance and decision-making
capabilities.

B. Overview of Financial Market

Financial markets, from the name itself, are a type of marketplace where different financial
assets such as stocks, bonds, and currency are bought and sold. Often referred to as "capital
markets" or even just "Wall Street," their primary goal is to assist companies raise money for
growth while giving investors a chance to profit on their investments.
Some special types of financial markets: Financial markets exist in various forms, each
specializing in different financial instruments.

Platforms like the NYSE and Nasdaq allow companies to list shares for trading.
Investors buy and sell stocks, enabling businesses to raise capital while
investors seek returns (Bodie et al., 2014).
Stock Markets

1
 These decentralized, electronic markets trade securities that do not meet
exchange criteria, often with less regulation and transparency (Hull, 2017).
OTC Markets

Specializing in short-term, low-risk investments like Treasury bills, they offer

safety and liquidity (Mishkin, 2015).
Money
Markets

Focused on debt instruments, this market allows governments and corporations

to issue bonds with fixed interest rates (Fabozzi, 2016).
Bond Markets

These markets trade contracts like futures and options based on other assets to
manage risk (Hull, 2017).
Derivatives
Markets

C. Trends of AI applications in the development of Banking Industry

The growing role of AI in banking is a key aspect of the sector's ongoing digital transformation.
With advancements accelerated by the COVID-19 pandemic, the demand for AI and machine
learning (ML) has surged. A 2021 PwC report reveals that over half of UK banks now recognize the
critical importance of AI. According to McKinsey, by 2019, approximately 60% of financial
services professionals had already implemented at least one AI solution.
Prominent AI technologies in banking include Robotic Process Automation (RPA) for handling
operational tasks, virtual assistants for enhancing customer service, and machine learning
algorithms for fraud detection and risk management.
Banks are increasingly leveraging AI across various operations to achieve the following:
1. Automating decision-making processes and offering intelligent recommendations.
2. Personalizing customer interactions through analyzing behavioral insights.
3. Providing smooth, consistent customer experiences across all platforms.
4. Integrating banking services with non-banking products for a comprehensive experience.
Through these applications, AI enhances efficiency, allows banks to analyze vast amounts of
customer data in real-time, and improves both predictive capabilities and operational
performance.

2
II. Applications of AI in Finance Markets

A. ANN
1. What are ANNs?
ANNs are computational models inspired by the human brain, consisting of interconnected
neurons that process data in complex, nonlinear ways to simulate cognitive functions. ANNs are
widely used in fields such as image and voice recognition, robotics, finance, and hydrology for
tasks including pattern classification, prediction, and optimization.
ANNs consist of three layers: an input layer (receiving the initial data), one or more hidden layers
(processing the data), and an output layer (providing the final prediction). The neurons are
connected by weighted links that adjust during learning. ANNs learn from data by adjusting the
weights of the connections between neurons. This learning is often achieved through a process
called "training," where the network is fed large amounts of data and gradually improves its
performance by minimizing errors.
2. How does ANNs work in Finance?
a. Stock Movement Prediction:
ANNs are utilized to forecast the direction of stock market indices and individual stock prices by
analyzing extensive historical data. They enhance prediction accuracy by considering various
inputs, such as historical prices, trading volumes, and economic indicators.
In one study, ANNs demonstrated a 94.08% prediction accuracy for NASDAQ stock movements
using an ANN with a three-layer structure (20-40-20 neurons in the hidden layers). These feed-
forward networks are favored due to their ability to generalize well and their straightforward
implementation.
b.Credit Card Customer Evaluation:
In the credit card industry, ANNs assist in deciding whether to approve or deny credit card
applications by analyzing historical data on customer behavior. By examining past experiences with
profitable and non-profitable customers, ANNs help predict which applicants are likely to be
profitable.
c. Loan Application Assessment:
ANNs are also used to evaluate loan applications by analyzing various factors to predict the
likelihood of successful loan repayment. These networks help financial institutions reduce the
default rate and increase the success rate of their loan portfolios.
Research showed that ANNs could predict loan recovery outcomes with 92.6% accuracy,
reducing Type I errors (approving risky loans) and Type II errors (rejecting safe loans)
3. Case Study
Assessing consumer loan risk is vital for minimizing defaults, requiring more than basic
financial metrics. A study made by professor Angelos P. Markopoulos in National Technology

3
 University Athens used Artificial Neural Networks (ANN) in MATLAB to analyze comprehensive
data, with high accuracy in predicting non-performing loans.
In this case study, an Artificial Neural Network (ANN) was employed to predict the risk of
consumer loans becoming non-performing (NPL) by analyzing both financial and personal data of
loan applicants. Using MATLAB's Neural Network Toolbox, data from 1,000 bank clients,
including 20 different attributes such as loan duration and personal status, were normalized and
processed for input into the ANN. The dataset was divided into training and testing sets, and various
configurations and training algorithms were tested to optimize the model's performance, with
Gradient Descent with Momentum and Adaptive Learning Rate Back-propagation (GDX) proving
most effective.
The results showed that the ANN accurately predicted the loan outcomes in 79.67% of the
cases, which increased to 81% when decision thresholds were adjusted for borderline cases. The
model's output, ranging between 1 and 2, allowed for a nuanced assessment of loan risks, helping to
identify high-risk loans that are more likely to become non-performing. This predictive capability
illustrates the potential of ANNs to enhance loan risk assessments by providing more detailed
insights than traditional methods.

Overall, this study demonstrates the practical application of ANNs in finance, particularly in
improving the accuracy and reliability of credit risk evaluations. By incorporating a wide range of
data inputs and using advanced training techniques, ANNs can serve as a valuable tool for banks to
better assess loan risks, ultimately aiding in more informed decision-making and reducing the
likelihood of loan defaults. (Makrygianni & Markopoulos, Angelos P, 2015)

B. Expert system
1. Basics of expert systems
An expert system is a type of computer software designed to emulate the expertise of a human
professional in a specific domain. It falls under the field of AI, which focuses on creating simulations
that replicate human problem-solving and learning processes. This system leverages the specialized,
often use knowledge of experts so that even individuals without expertise can benefit from it. The
core components of an expert system are the inference engine and the knowledge base. The inference
engine integrates user input with the knowledge base to generate and apply rules, thereby creating
new insights from the data. It is also capable of explaining and diagnosing issues. Meanwhile,
developing a knowledge base involves gathering information and mapping out the relationships
between data points, typically using a set of IF-THEN rules. Each component of an expert system
plays a role in the decision-making process.
 Fact Database: This contains all the information submitted by users.
4
  Knowledge Base: More than just a repository of data, it encompasses the relationships,
outcomes, and predictions derived from the data. The knowledge base is managed through
facts and rules, and knowledge engineering is employed to extract expert knowledge and
convert it into these facts and rules.
 Inference Engine: This component connects the knowledge bases with the factual data to
generate variables and draw conclusions based on them.
 Interpretation Mechanism: Crucial for expert systems, this stage explains how the system
arrived at a decision and clarifies the reasons behind particular questions. Given that the
system operates on rules, it can detail the rules applied during the inference process, helping
users assess and verify the reliability of the decision.
 User Interface: This component enables users to input their data and receive answers and
results to their queries.
Unfortunately, information about such systems is generally limited, since disclosure of
successful approaches by the financial operators could lead to the loss of competitive advantage,
and large sums of money. As a general point, financial operators today tend to prefer neural-
networks for real-time forecasting, while expert systems now tend to be used more in other financial
fields, where the outcome of the system must be a clear decision - i.e. validating user's credit-card
accesses. Expert systems are used in accounting, auditing, decisions in insurance companies, etc.

Figure 1: General architecture of expert systems

2. Applications of expert systems in financial sector
 Expert systems for financial analysis of firms. An advantage of expert systems is that
besides quantitative ratios such as profitability, total profit, short term and long-term debt
and other, they can treat qualitative ratios also: position at the market, organization of
personnel, business reputation, marketing flexibility, etc. Successful financial analysis
determines the firm’s status, that is, a level of risks for a possible investment or a credit
approval.
 Expert systems for analyzing the causes of successful or unsuccessful business
development (Apte 1989; Pinson, 1992; Ruparel & Srinivasan, 1992; Chinn & Madey,

5
 1997). Because of the possibility to draw a conclusion about a certain business development
or a specific project by looking backward or forward, and also the ability to follow data that
can change over time, it is possible to find causes of the temporary state of affairs in the
business/project and predict its future.
 Expert systems for market analysis (Chan, Dillon & Saw 1989; Dhananjayan, Raman, &
Sarukesi 1989; Smith, McDuffie & Flory 1991). If a product has already been produced
before by a company, then its sale can be analyzed by an expert system. The system can take
into account different factors that can possibly decrease the sale (high price, low quality, bad
commercial, stiff competition, etc.), and then on the bases of the analysis the company
should decide on further steps (to improve quality, to improve production, or to start
producing a new product).
 Expert systems for acquiring knowledge in a subfield of finance (Boer & Livnat 1990;
Brown & Wensley 1995; Hartvigsen 1992). These expert systems are widely used in
educating managers and other financial experts. Besides, this kind of knowledge can
advance and be improved rapidly, so the knowledge bases of such expert systems are
suitable means for its conservation, improvement and reusability.
3. Case study (the use of INVEX system in investment management)
The INVEX expert system helps the project analyst and investment decision-maker to determine
whether a project is acceptable and, if it is, whether it is the best alternative, and to calculate the
extent of the decision sensitivity to certain critical assumptions.
During a consultation, INVEX first asks about a customer's preferences and intentions, then
builds up a customer profile, where the information asked from customers depends heavily on their
intentions and the course of the consultation.
Once all input data is prepared in Excel, INVEX follows these steps: It categorizes investments
into five groups based on dynamic parameters. Investments labeled VERY GOOD proceed to
multicriteria decision-making (MCDM), while those labeled VERY BAD are rejected. For
investments classified as GOOD, MEDIUM, and BAD, a group-specific sensitivity analysis
(GSSA) is conducted, and users decide whether to accept or reject them. This process reduces the
pool for MCDM by eliminating poor choices. If specified, a risk analysis is performed on accepted
investments, leading to an optimal investment combination. Skipping the initial categorization may
result in selecting investments that, while optimal, might not be acceptable to the investor. The final
selection does not necessarily include all top-ranked investments..

6
 Figure 2: Divisions of investments in INVEX
The financial viability of the project can be assessed, and alternative projects can be
categorized into five groups based on their level of desirability.

GROUPS v r p c

A (VERY GOOD) >1 >25% <p0 <c0

B (GOOD) >0 >10% and <25% <p0 <c0

C1 (MEDIUM <0 >6% and <10% >p0 <c0

C2 (MEDIUM) >0 >10% and <25% >>p0 >c0

D (BAD) <-1 <7% >>p0 >c0

E (VERY BAD) <-1 <4% >>p0 >>c0

Figure 3: The groups projects in INVEX

Legend: v, relative net present value of investment; r. return on investment; p. payback period: c. period of
achieving the critical breakeven point; po, reference payback period: co, reference period of achieving the critical
point.

INVEX assumes that the user will definitely accept the investments from the group VERY
GOOD and reject the investments from the group VERY BAD. For the investments from group
GOOD, MEDIUM and BAD, where the user is not sure whether to accept or reject a particular
investment, the system performs a group-specific sensitivity analysis which gives some additional
information that can help the users to make up their minds. Note that if the total preorder is used, it

7
is not possible to give any additional information to the users that can help them to decide whether
to accept or reject a particular investment.

C. Hybrid Intelligence in Finance

Hybrid intelligence integrates AI technologies with human expertise to optimize decision-
making in finance. It is increasingly used across key financial strategies, including portfolio
management, risk management, and stock market prediction.
1. Portfolio Management
Portfolio management is a critical financial strategy that involves balancing risk and reward to
achieve optimal returns. Hybrid intelligence enhances this process by enabling AI to handle large
volumes of data, identify market trends, and suggest asset allocation strategies. AI models can
continuously process historical and real-time data to detect patterns that may indicate future market
movements. This data-driven insight allows portfolio managers to make informed decisions.
However, portfolio management still requires human oversight, especially when considering
external factors like political events or regulatory changes that may not be fully captured by AI
algorithms. Hybrid intelligence ensures a balance between data-driven recommendations and
qualitative insights, leading to more comprehensive portfolio management decisions.
2. Risk Management
In financial markets, managing risk is essential to minimizing potential losses. Hybrid
intelligence systems monitor market conditions in real-time, identifying potential risks such as
credit defaults, market volatility, or liquidity issues, with human experts interpreting complex
scenarios. AI models excel at processing historical risk data and predicting future risks based on
current market conditions.
Despite the power of AI, human judgment remains crucial in interpreting risks that cannot be
fully quantified, such as geopolitical instability or sudden economic shifts. Hybrid systems combine
AI’s predictive power with human decision-making to ensure that risk management strategies are
both proactive and adaptable.
3. Stock Market Prediction
Stock market prediction is one of the most challenging areas of finance, given the market’s
volatility and the sheer number of factors influencing stock prices. Hybrid intelligence has proven
to be particularly effective here. AI models can process large datasets, including stock prices,
trading volumes, and market sentiment, to predict future movements. These models are particularly
useful in identifying trends and executing high-frequency trades.
However, stock market behavior is also influenced by factors that AI may not easily quantify,
such as sudden regulatory changes or shifts in investor sentiment. Human expertise adds value by
assessing these qualitative factors, ensuring that predictions are not solely based on historical data
but also consider the broader market context. The combination of AI-driven insights and human
experience results in more accurate stock market predictions.
4. Application of Hybrid Intelligence at BlackRock

8
 BlackRock, the world’s largest asset manager, has demonstrated its ability to leverage cutting-
edge technologies to maintain its leadership of the financial industry. As of June 2024, BlackRock
manages an impressive $10,646 billion in assets under management (AUM), having grown from
$53 billion in 1994, representing a compound annual growth rate (CAGR) of 20%. Central to its
growth and operational efficiency is the use of hybrid intelligence, which combines artificial
intelligence (AI) with human expertise to optimize portfolio management and risk mitigation
strategies. This integration is primarily executed through BlackRock’s proprietary platform,
Aladdin.

Source: BlackRock Annual Reports

Aladdin, which stands for Asset, Liability, Debt, and Derivative Investment Network, is the
backbone of BlackRock’s risk management and investment processes. The platform employs AI-
driven analytics to process vast quantities of financial data from global markets, offering real-time
insights into asset allocation, market trends, and risk assessments. By utilizing machine learning
algorithms, Aladdin can quickly identify patterns in the data, flag potential risks, and suggest
portfolio adjustments, which are then reviewed by human experts.. These AI-driven insights
provide BlackRock with a competitive edge in managing its massive portfolios more efficiently
than human analysts alone could achieve (Soltani, 2024).
However, Aladdin’s success lies not only in its AI capabilities but in the human expertise that
guides its use. While AI processes vast datasets and performs predictive analytics, BlackRock’s
portfolio managers and risk officers evaluate these insights to account for qualitative factors that AI
may not fully capture, such as geopolitical events or unexpected regulatory changes. This
combination ensures that investment decisions are not only data-driven but also adaptable to real-
world complexities, allowing BlackRock to make well-rounded, informed decisions.

III. Case study - Vietcombank

1. Overview
Founded 60 years ago, Vietcombank is a leading Vietnamese bank with over 600 branches and
nearly 23,000 employees. Ranked among The Banker’s Top 500 Global Banks and the Top 30
Asia-Pacific banks.

9
2. Application of AI in Vietcombank
Vietcombank has implemented a hybrid system for credit scoring that combines AI-driven
algorithms with human expertise. The AI component analyzes large volumes of data, including
credit history, transaction patterns, and social factors, to provide an initial credit score. Human
experts then review these scores, considering additional context and subjective factors that the AI
might not fully capture.
Vietcombank's integration of AI into its credit scoring system aims to significantly enhance the
precision of credit assessments, thereby minimizing default risks. By leveraging advanced
algorithms, the bank accelerates the credit approval process while upholding rigorous risk
assessment standards, ultimately increasing efficiency. This modernization also contributes to a
more transparent and expedited decision-making experience for customers. Despite the reliance on
AI, Vietcombank ensures that human judgment and oversight remain crucial components of the
process, maintaining a balanced approach that combines technological innovation with essential
human expertise.

3. Human Oversight
Despite the critical role of AI in the credit scoring process, Vietcombank ensures that all credit
decisions undergo human review. Credit analysts are tasked with examining cases where AI-
generated scores are unusually high or low, or where the applicant's profile presents complexities
that require additional scrutiny. Experienced credit officers apply their expert judgment to consider
qualitative factors that AI might not fully capture, such as recent economic changes or unique
individual circumstances, thereby adding a layer of nuanced evaluation to the automated process.

4. Results
AI-powered credit scoring at Vietcombank has led to several notable improvements. AI has
improved risk predictions, reduces loan defaults and refines the identification of high-risk
applicants. The automation of credit scoring has sped up application processing times, resulting in
faster credit approvals and more efficient use of credit officers' time, allowing them to handle more
complex cases. Customers benefit from quicker decisions and a more transparent application
process, with AI providing clear criteria and data-driven explanations. Despite the advancements in
automation, human oversight remains integral, ensuring that the process is adaptable to exceptional
cases and mitigating potential biases to ensure fair treatment of all applicants.
After 6 months of implementation, VCB Digibot has handled 88.5% of customer requests, with
only 11.5% requiring human advisors. Nearly two million requests have been successfully
managed. VCB Digibot now manages over 100,000 queries per month, greatly improving
response times and customer satisfaction. AI-based fraud detection systems have reduced fraudulent
transactions by 30% over the past six months, enhancing security and minimizing financial losses.
Additionally, AI has boosted operational efficiency by 20%, primarily through streamlining
workflows and decision-making processes streamlining various banking processes and reducing the
time required for decision-making by 15%. These advancements have solidified Vietcombank’s
position as a leading financial institution and contributed to its ongoing success in the banking
sector.

10
Conclusion
Vietcombank's approach to combining AI-powered credit scoring with human oversight
represents a forward-thinking strategy that enhances both the efficiency and accuracy of credit
assessments. By leveraging advanced technology while maintaining critical human judgment,
Vietcombank has modernized its credit scoring process, improving risk management, customer
satisfaction, and operational flexibility.

IV. Challenges & risks of AI in Finance

Challenges of AI in Finance
1. Regulatory Compliance and Transparency
AI systems are often opaque, making it difficult for both institutions and regulators to understand
and monitor how decisions are made. In finance, compliance is critical, and regulators need clear
audit trails for AI-driven decisions, particularly in areas like lending, trading, and fraud detection.
According to a 2022 PwC survey, 60% of financial services executives believe that regulatory
uncertainty is the primary challenge in adopting AI systems. Moreover, the "black box" problem,
where the internal workings of AI models are difficult to explain, hampers trust and compliance
with financial regulations like MiFID II and Basel III.
2. Bias and Ethical Considerations
AI systems can inherit biases from the data they are trained on. For example, in credit scoring or
loan approval processes, historical biases could lead to discrimination against minorities or low-
income individuals. Ethical AI use is now a major concern, and financial institutions are
increasingly focused on creating AI systems that are transparent and fair. Companies like JP
Morgan and Goldman Sachs are investing in frameworks to minimize bias, but there is still
significant progress to be made.
3. Data Quality
Data quality is a critical challenge for the implementation of artificial intelligence (AI) in finance.
The accuracy and effectiveness of AI models heavily depend on the quality of data used to train
them. Therefore, poor data quality can lead to inaccurate results and unreliable predictions.
According to a recent study by Deloitte, data quality is one of the most significant challenges
facing organizations looking to implement AI in finance. The study found that data quality issues
are among the top three reasons why AI projects fail to meet their objectives. The study also
revealed that 57% of respondents considered poor data quality as a barrier to successful AI
implementation.

Risks of AI in Finance
1. Data Security and Privacy Concerns

11
 AI systems rely on large datasets, often containing sensitive financial information. This
dependence makes AI systems particularly vulnerable to cyberattacks and data breaches. According
to IBM’s "Cost of a Data Breach Report" (2022), the average cost of a financial sector breach was
$5.97 million—significantly higher than the global average of $4.35 million.
Additionally, the growing use of AI raises concerns about compliance with data protection
regulations like GDPR in Europe and CCPA in California. A breach not only damages consumer
trust but also can result in heavy fines. AI systems must be designed to safeguard this data while
maintaining compliance with complex regulatory frameworks.
2. Systemic Risks and Market Manipulation
The 2010 Flash Crash serves as a key example of these risks, where the Dow Jones Industrial
Average dropped nearly 1,000 points within minutes, largely due to high-frequency trading
algorithms. The CFTC and SEC identified that a large sell program executed by an algorithm led to
cascading effects across markets, demonstrating the systemic risks posed by such automated trading
activities.

3. Fraud and Misuse of AI

A report by PYMNTS highlighted that AI-driven fraud detection in financial institutions is
facing challenges from sophisticated cybercriminals. 43% of financial institutions reported an
increase in fraud, with digital payment methods like digital wallets and ACH payments being highly
targeted.
According to NICE Actimize 2022 Fraud Insights report, banking and payments transactions
using mobile devices have increased substantially. But unfortunately, the rise in mobile usage has
been accompanied by increased instances of fraud – 61% of attempted fraud attacks through mobile
apps are Account Takeovers (ATOs), the data showed.
Fraudsters are exploiting the prevalence of mobile usage to target and leverage older devices.
The report also showed that cell phones using older operating systems or made before 2016 have
three times as many fraud attempts as newer devices or operating systems. Nearly half (46.9%) of
attempted fraud stemmed from card-not-present transactions across payment channels. As a result,
online transactions presented a growing focus in the threat landscape.

12
4. Job Displacement and Workforce Challenges
Automation driven by AI is reshaping the job landscape in finance. Artificial intelligence, a key
driver of potential algorithmic displacement, is adopted by nearly 75% of surveyed companies and
is projected to lead to high churn – with 50% of organizations expecting it to create job growth and
25% expecting it to create job losses.
However, new roles will emerge that require advanced data science, machine learning, and AI
expertise. The challenge lies in reskilling the existing workforce to adapt to these new roles.
Financial institutions are investing heavily in training programs, but the transition could take years
and lead to significant economic displacement in the interim.
Conclusion
AI’s potential to revolutionize the financial industry is evident in its ability to enhance
decision-making, streamline operations, and improve customer experiences. However, these
benefits come with significant challenges and risks that must be carefully managed. Financial
institutions must prioritize transparency, data security, and ethical standards while ensuring that
they comply with evolving regulations. With careful implementation and oversight, AI can lead
to a more efficient and inclusive financial system, but vigilance will be crucial in mitigating its
inherent risks.

V. Future trend and development of artificial technology in the financial market.

1. The Use Of Predictive Analytics
Predictive analytics is being used in the financial services industry to identify potential risks,
optimize lending and investment decisions and improve customer targeting. A major use case for
predictive analytics within investment firms is developing predictive models for algorithmic trading
and then executing market-making decisions within milliseconds. These models typically analyze
vast amounts of historical data, as well as real-time market data, to identify patterns and predict
future movements in the stock market. Other majors use cases include intelligent credit scoring and
proactive customer support before major issues occur.
Hedge funds like Two Sigma and Citadel already use AI-powered trading algorithms to predict
stock price movements and execute trades. These systems continuously analyze market signals and
data from various sources, helping traders make decisions in milliseconds, something human traders
cannot achieve.
2. AI in Banking and Customer Service
AI's impact on the banking sector is equally transformative. An increasing number of financial
companies use various different technologies to offer digital online services that have traditionally
been provided by mainstays of the financial industry. Many financial services firms are using AI to
detect fraud, predict cash-flow events, conduct cost and benefit analysis, as well as for account
creation and goal setup. Other uses include recommendations for investing and retirement planning,
communication between users for mutual investments, and trading and investing in stocks, bonds.
Continuous 24/7 accessibility, automated rebalancing, and monitoring differentiate robot-
advisors from traditional investment advisory services. Customers can access their accounts via
13
user-friendly websites or smartphone applications and make adjustments to their portfolios any time
of the day and recalibrate their investments.
Bank of America has developed an AI-powered virtual assistant named Erica, which helps
customers with tasks such as tracking spending, finding transactions, and offering financial advice.
By analyzing users' financial behaviors, Erica provides personalized insights and helps customers
make better financial decisions.
3. The Growing Use Of Blockchain Technology
Blockchain technology is also becoming increasingly popular in the financial services industry
as a way to improve security and transparency. Banks are exploring the use of blockchain for
various use cases such as digital identity, trade finance and cross-border payments.
Blockchain consists of a series of "blocks," each containing a list of transactions. These blocks
are cryptographically linked to one another in a "chain," forming a complete and unalterable history
of transactions. Each participant in the network holds a copy of the blockchain, making the system
decentralized. When a new transaction is added, it must be validated by a consensus mechanism
(e.g., Proof of Work or Proof of Stake) before it can be added to the chain. Once validated, the
transaction becomes permanent and cannot be altered.
Blockchain's most famous application is in the world of cryptocurrencies, such as Bitcoin and
Ethereum. However, its impact on the broader financial system goes far beyond digital currencies.
Blockchain enables faster, cheaper cross-border payments, secure and transparent transactions, and
the potential to create decentralized financial systems.
This report highlights how Artificial Intelligence (AI)—the ability of machines to simulate
human-like processes—is transforming financial markets. John McCarthy first introduced the
concept at the 1956 Dartmouth Conference, defining AI as the science of creating intelligent
machines. The Financial Stability Board similarly defines it as the development of systems capable
of tasks requiring human intelligence. AI automates learning and discovery through data, utilizing
advanced algorithms to analyze information and make decisions with minimal human input, based
on three components: Data, Algorithm, and Learning Mechanism.
Conclusion
In financial markets, AI plays a pivotal role, particularly in banking. Financial institutions
are leveraging AI to streamline decision-making, personalize customer interactions, and enhance
operational efficiency. AI allows banks to conduct real-time data analysis, improving predictive
accuracy and customer experience.

14
 Conclusion
In summary, this report has illuminated the profound ways in which Artificial Intelligence (AI)
- the capability of machines to simulate human-like processes, such as learning, adapting to new
information, and performing complex tasks that is reshaping the financial markets.
In financial markets, which encompass a wide range of platforms from Stock Markets to
Cryptocurrency Markets where various assets like stocks, bonds, and currencies are traded, AI has
proven its pivotal role, particularly in the banking sector. Banks are increasingly leveraging AI to
streamline decision-making, provide intelligent recommendations, and personalize customer
interactions through behavioral analysis.
Furthermore, the report has highlighted the substantial benefits AI brings to the industry of
Artificial Neural Networks, Expert Systems, and Hybrid Intelligence
 Artificial Neural Networks (ANNs)
o Improved prediction accuracy: ANNs enhance stock movement prediction
accuracy by analyzing various inputs, such as historical prices and economic
indicators, as demonstrated by a 94.08% prediction accuracy for NASDAQ stock
movements.
o Enhanced customer evaluation: ANNs improve credit card customer evaluation by
analyzing historical data, allowing banks to better predict which applicants will be
profitable, minimizing error rates and optimizing credit card issuance strategies.
o More accurate loan assessments: ANNs help financial institutions evaluate loan
applications with higher precision, reducing the risk of approving risky loans or
rejecting safe ones, with a demonstrated 92.6% accuracy in predicting loan
repayment outcomes.
 Expert Systems
o Handling both quantitative and qualitative data: Expert systems can analyze not
only financial ratios like profitability, profit, and debt but also qualitative factors
such as market position, personnel organization, business reputation, and marketing
flexibility.
o Comprehensive financial analysis: They help in determining a firm's risk level for
investment or credit approval by providing in-depth analysis.
o Identifying causes of business success or failure: Expert systems can analyze
business development retrospectively or prospectively, helping to identify reasons
for the current state and predict future performance.
o Market analysis: They can assess product sales by considering factors such as price,
quality, marketing, and competition, which helps companies make strategic decisions
(e.g., improving quality or shifting focus).
 o Knowledge acquisition and education: Expert systems are useful for educating
financial professionals and conserving, improving, and reusing financial knowledge,
as their knowledge bases evolve rapidly.
o Automation and standardization: Expert systems like Port-Man can streamline the
financial advisory process by offering personalized investment advice based on
predefined criteria, ensuring consistency in financial consultant advice.
 Hybrid Intelligence
o Enhanced portfolio management: Hybrid intelligence improves portfolio
management by allowing AI to analyze large volumes of data, detect market trends,
and provide asset allocation strategies. At the same time, human oversight adds
qualitative insights, such as political or regulatory factors, ensuring well-rounded
decision-making.
o Improved risk management: Hybrid intelligence strengthens risk management by
combining AI's ability to predict risks based on historical and real-time data with
human judgment, which is crucial for assessing risks related to unpredictable factors
like geopolitical events or sudden economic changes.
o More accurate stock market prediction: Hybrid intelligence excels in stock market
prediction by leveraging AI to process vast datasets and identify trends. Human
expertise complements this by interpreting non-quantifiable influences, such as
regulatory changes or investor sentiment, resulting in more reliable predictions.
Artificial Intelligence plays a pivotal role in Vietnam's financial sector, particularly in banking.
Vietcombank has adopted a hybrid credit scoring system that blends AI-driven algorithms with
human expertise. Despite the heavy reliance on AI, human oversight remains essential for
addressing complex cases and mitigating potential biases. By integrating AI with human oversight,
Vietcombank has not only modernized its credit scoring process but also solidified its position as a
leading financial institution, driving its ongoing success.
Despite playing a pivotal role in Vietnam’s financial sector, the adoption of AI faces several
significant challenges. Regulatory compliance is hindered by the lack of transparency in AI
systems, often referred to as the "black box" problem, which complicates oversight. Additionally,
AI can perpetuate biases from historical data, raising ethical concerns, while the need for high-
quality data is crucial to ensure accurate predictions. AI also brings risks, such as data security
vulnerabilities, systemic threats from high-frequency trading algorithms, increased fraud in digital
banking, and potential job displacement due to automation
In terms of the future, AI in finance is set to revolutionize the industry, driven by innovations
like generative AI, predictive analytics, and blockchain technology. AI is enhancing customer
service, risk management, and capital markets, while predictive analytics is optimizing lending,
investment decisions, and fraud detection.
In conclusion, Artificial Intelligence (AI) is revolutionizing the financial industry by enhancing
operational efficiency, accuracy, and customer experience. As financial institutions increasingly
integrate AI into their operations, they are able to make more informed decisions, reduce errors, and
provide faster, more tailored solutions to their clients. The continued advancement of AI promises
to drive further innovation in finance, making the industry more dynamic, efficient, and responsive
to evolving market needs.
 References

1. Hussain, M. (2024). Council Post: The Future Of Data And AI In The Financial
Services Industry. Forbes. [online] 12 Aug. Available at:
https://www.forbes.com/councils/forbestechcouncil/2023/02/27/the-future-of-data-and-
ai-in-the-financial-services-industry/.
2. Chlouverakis, K. (2024). How artificial intelligence is reshaping the financial services
industry. [online] www.ey.com. Available at: https://www.ey.com/en_gr/financial-
services/how-artificial-intelligence-is-reshaping-the-financial-services-industry.
‌3. Worldfinance.com. (2023). The future role of AI in finance. [online] Available at:
https://www.worldfinance.com/markets/the-future-role-of-ai-in-finan [Accessed 23 Sep. 2024].
4. V ‌ ietnam+ (VietnamPlus (2023). Vietnam develops AI technology industry. [online] Vietnam+
(VietnamPlus). Available at: https://en.vietnamplus.vn/vietnam-develops-ai-technology-industry
[Accessed 23 Sep. 2024].
5. ‌Tran, M.C. (2023). AI in Vietnam: A Quick Guide. [online] Vietnam Briefing News. Available
at: https://www.vietnam-briefing.com/news/ai-in-vietnam-opportunities-and-challenges.html/.
6. ‌Barnes, M. (2024). Survey Finds AI, Tech Popular in Financial Services in Vietnam. [online]
Vietnam Briefing News. Available at: https://www.vietnam-briefing.com/news/survey-finds-ai-
tech-popular-in-financial-services-in-vietnam.html/.
7. ‌vietnamnews.vn. (2024). Harnessing AI for Việt Nam’s economic future. [online] Available at:
https://vietnamnews.vn/economy/1662075/harnessing-ai-for-viet-nam-s-economic-future.html
[Accessed 23 Sep. 2024].
8. ‌News, V.N. (2023). AI transforming banking, financial sectors in Vietnam. [online] Asia News
Network. Available at: https://asianews.network/ai-transforming-banking-financial-sectors-in-
vietnam/.
9. ‌Lapphuong.com. (2022). Các chỉ số tài chính cơ bản. [online] Available at:
https://digital.lapphuong.com/vcb/ar2022/final/vn/cac-chi-so-tai-chinh-co-ban.html [Accessed
23 Sep. 2024].
10. ‌Investopedia. (2024). Financial Markets: Role in the Economy, Importance, Types, and
Examples. [online] Available at: https://www.investopedia.com/terms/f/financial-
market.asp#:~:text=Financial%20markets%20refer%20broadly%20to%20any%20marketplace
%20where%20the%20trading [Accessed 23 Sep. 2024].
11. ‌ITChronicles. (n.d.). Artificial Intelligence - A Comprehensive Overview of All Things AI.
[online] Available at: https://itchronicles.com/what-is-artificial-intelligence/.
12. ‌CFI Team (2023). Financial Markets. [online] Corporate Finance Institute. Available at:
https://corporatefinanceinstitute.com/resources/career-map/sell-side/capital-markets/financial-
markets/.
13. ‌tapchinganhang.gov.vn. (n.d.). Xu hướng ứng dụng trí tuệ nhân tạo trong sự phát triển của
ngành Ngân hàng. [online] Available at: https://tapchinganhang.gov.vn/xu-huong-ung-dung-tri-
tue-nhan-tao-trong-su-phat-trien-cua-nganh-ngan-hang.htm.
14. ‌Deloitte (2023). How Artificial Intelligence Is Transforming the Financial Services Industry.
[online] www.deloitte.com. Available at: https://www.deloitte.com/ng/en/services/risk-
advisory/services/how-artificial-intelligence-is-transforming-the-financial-services-
industry.html.
15. ‌