Influence of Customer Perception Factors on AI-Enabled Customer Experience in the Ecuadorian Banking Environment

Abstract

This study reviews the relationship between customer perception factors and AI-enabled customer experience in the Ecuadorian banking industry. The study employs a self-designed online questionnaire with five factors for customer perception (convenience in use, personalization, trust, customer loyalty, and customer satisfaction) and two categories for AI-enabled customer experience (AI-hedonic customer experience and AI-recognition customer service). The final valid dataset consisted of 226 questionnaires. The data analysis and the hypotheses tests were conducted using SPSS 26 and structural equation modeling, respectively. The main findings displayed that all five customer perception factors (individual and joint effect) have a positive and significant effect (at least at the 5% level) on AI-enabled customer experience, AI-hedonic customer experience, and AI-recognition customer service in the Ecuadorian banking industry. Study results are aligned with previous findings from other countries, particularly the banking environment in the United Kingdom, Canada, Nigeria, and Vietnam. The AI techniques involved in the financial sector increase the valuation of customer experience due to AI algorithms recollecting, processing, and analyzing customer behavior. This study contributes a complete statistical and econometric model for determinants of AI-enabled customer experience. The main limitations of the study are that, in the analysis of the most demanded AI financial services, not all services and products are included and the inexistence of a customer perception index. For upcoming research, the authors recommend performing a longitudinal study using quantitative data to measure the effect of AI-enabled customer experience on the Ecuadorian banks’ performance.

1. Introduction

Artificial intelligence (AI) comprises an extensive field of science including areas such as psychology, computer science, linguistics, and philosophy. It generates benefits for the global economy, especially in economic sectors that provide services to customers, provided AI might improve the quality of services and products. Specifically, USD 1 trillion each year might be created by the implementation of AI in the global banking industry [1,2]. The allure of AI-based financial institutions is based on lower operating costs, better client approximation, higher revenues, and innovation in financial products and services. Nowadays, the customers’ preferences and perceptions have been captured by banking industry, especially due to most financial operations being digitized using AI algorithms. Therefore, it is imperative to understand the level of customer knowledge about the implementation and functionality of AI technologies. However, despite the potential interactions between customer perception factors and AI-enabled customer experience, little research has been conducted on the role of client perception factors in AI-enabled experiences in the Ecuadorian banking industry. Similar studies have been developed in other countries in Asia, Europe, or North America, and the results showed a positive relationship between customer perception factors and AI-enabled customer experience. To know the type of relationship in the Ecuadorian case can help financial institutions to create a competitive advantage by implementing AI in their services.

Customer perception factors include thoughts and impressions about the financial entity. This information is collected, processed, and analyzed to comprehend the needs, demands, and purchase performance of customers to create a long-term bank–client connection based on the knowledge of clients. Satisfied customers will continue doing business with a specific bank, which is transformed into lower costs for the bank, creation of value [3], incorporation of new and creative AI strategies in all processes, and sustainable competitive advantage. Banks offer undifferentiated financial products and services, and thus, factors involved in client perception convert the dominant competitive preference between parties, motivating positively the development and incorporation of digital banking products and services with AI.

Gaining customers may be recognized as the fundamental aim of a firm, and thus, the objective for a firm is to maintain and improve the business relationship with customers by building a strong long-term link between financial entities and customers. Therefore, entities might reduce the breach between expectations and real financial products and services, where AI plays an important role as it has changed the traditional methods of co-creating value in firms, products, and services [4]. Moreover, the application of AI implies improvements in marketing practices and services, which can rapidly extract the preferences of economic actors [5]. With AI, intuitive and empathic tasks can be emulated by designing cognitive neural networks to identify customer personalities and interact with humans to create value [6,7]. However, few studies have been explored with this concern.

The research question of this study refers to identifying the factors that influence the perception of clients in the banking industry, especially their viewpoint when they access the virtual/mobile bank app, which is also created using AI algorithms, and how their customer perception is modified by the AI-enabled user experience. Aligned with the research question, the main objective of this study is to recognize the relationship between customer perception factors and AI-enabled customer experience in the Ecuadorian banking industry using five criteria to measure the customer perception: convenience in use, personalization, trust, customer loyalty, and customer satisfaction, while AI-enabled experience is measured by AI-hedonic customer experience and AI-recognition customer service (more details are given in the next section). Similar to previous investigations in other countries, the findings revealed that AI-enabled customer experience is affected positively (at least at the 5% level of significance) by all five customer perception factors using individual and joint effects.

This study provides significant insights for bank managers given customer perception and opinion of AI financial services offered by banks positively influence AI-enabled customer experience. This research provides an exhaustive view of the crucial customer perception factors that persuade AI-enabled customer experience. In addition, this study will instruct the banking and finance industry’s representatives and managers to design and implement a strategic plan using AI technology in their financial products and services. Moreover, it provides practical evidence of the customer-preferred factors when they use virtual/mobile financial apps and access digital services in the banking environment.

The rest of the paper is organized as follows. Section 2 displays a literature review and hypotheses development. Section 3 illustrates the empirical design. Section 4 presents the empirical findings. Section 5 establishes the discussion part. Section 6 exhibits conclusions and recommendations.

2. Literature Review and Hypotheses Development

2.1. AI in the Banking Environment

AI is defined as “the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages” [8]. AI combines algorithms that imitate the processes of human intelligence with machines that tend to emulate human behavior and cognitive functions. AI has been incorporated into several corporate processes where the volume of data far exceeds the capacity of people to acquit, interpret, and make decisions using voluminous databases. Therefore, AI involves the analysis of data and the solution for a certain task or problem in the shortest possible time.

One of the economic sectors that has benefited from the implementation of AI is the financial and banking industry. The fundamental impact of AI in the banking industry allows the increase in operational efficiency, optimization of resources, and growth of the firm’s profitability through the application of methodologies based on the combination of information and technology. Specifically, the financial and banking sectors have been impacted by AI in the following aspects [2,9,10].

Credit evaluation (credit score), risk management, cost saving, and automation: AI designs faster and more accurate evaluation of potential clients with the use of fewer resources, allowing a non-subjective identification of high-risk clients through the application of personalized learning models. Therefore, the credit financial planning of each bank/institution is more precise using models with the interaction of multiple variables, which are adjusted in real-time. Moreover, the operating cost of banks decreases given the reduction in personnel and paper costs involved in the credit score evaluation. Finally, the banking industry employs digital machines in different functional areas given the daily business volume in banks, which reduces the work stress and mathematical error in financial operations.

Prevention of credit card fraud and money laundering: Given the increase in electronic commerce and online transactions, AI algorithms analyze customer behavior, location of the user, and purchasing habits, to establish mechanisms and schemes of security in all purchases. Moreover, through the recognition of patterns, financial crimes can be prevented by the identification of suspicious financial activities and unusual movements of money.

Identification of investment opportunities: AI provides monitoring advantages to design intelligent investment systems to control structured information (databases) and unstructured information (text format). Therefore, investors might build and manage solid portfolios with fewer resources, time, and effort.

Banking and personalized attention (customer-centric): New products and services are offered through intelligence messaging robots, which provide immediate solutions and offer personalized financial advice, achieving faster decisions and transactions. Chatbot technology and virtual/mobile financial apps are the most useful AI tools in the banking industry, which interact with clients using preprogrammed queries that solve instant problems with effective communication.

However, the financial and banking sectors need to consider the biases of AI-enabled systems [11] and minimize their effects, for example, prevent the discrimination of minority groups when using credit score AI models [12]. Even when technology is neutral, human influence, datasets, or evaluation methods can cause biases in AI systems. Many efforts are being deployed to reduce this effect such as careful selection and governance of the data used in the AI models or diverse teams to reduce personal biases. Moreover, to maintain the principles of ethical AI (e.g., safety, explainability, security, fairness, human agency oversight, and privacy), governments are already taking action. For instance, the European Commission and Australia have created legal frameworks on AI ethics to guide organizations to develop and use AI in a responsible manner. These legal frameworks will help organizations to implement concrete actions in their daily operations and address more ethical and responsible AI solutions.

2.2. AI-Enabled Customer Experience

Customer experience in the banking environment refers to the adoption and acceptance of digital banking services. Even when customers’ predilections change between generations, customers will demand fast replies with custom-made attention and content. The current technology and the use of AI can help the banking sector to address useful customer information. For instance, by the scrutiny of customers’ surveys, emails, and social media interactions, it is possible to analyze customers’ sentiments and predict their emotions and feelings (what they like and dislike), and their interactions with the brand [13]. AI technologies also include machine learning and natural-language understanding and processing, which provide qualified feedback on customer sentiments, and retailers might improve the customers’ experience by promoting the firm’s competitive advantages [14]. For customers, this will improve their experience and meet their expectations [15]. Previous studies identified at least four elements that integrate customer experiences [13,15]: (i) cognitive sense (e.g., functionality, speed, and availability of service), (ii) emotional experience (e.g., positive or negative feelings generated by the service), (iii) physical and sensory contact (e.g., lighting, layout, signage, technology-related features, clear design, and friendly-user interface), and (iv) social factors (e.g., customer’s wider social network, customer’s social, and mental identity) [16,17,18]. AI technologies boost customer experiences by consuming data such as profiles [14], preferences, and past experiences of customers, to improve the customer–manager communication relationship and anticipate future consumption tendencies. Therefore, as mentioned previously, the interaction and integration of AI-enabled experience might increase the quality of the customer experience. AI-enabled customer experience includes hedonic and recognition customer experience and service.

2.2.1. AI-Hedonic Customer Experience

It means that clients may be prepared to prioritize hedonic utility with the customer’s strong self-relevant values. Hedonic consumption focuses on the most relevant aspects related to the acquisition experience, which are based on the subjectivity of the individual (e.g., sensations, memory, emotions, and fantasy). Therefore, consumption with hedonic characteristics has a diminutive weight of rationality, their benefits are not linked to the function, and they are connected with the satisfaction of emotions and ideas. Hedonic customer experience provides raw material to build, maintain, and motivate dialogues and relationships between customers and brands/products. Hedonic consumption experiences (e.g., beverages, perfumes, trips) facilitate the relationship and decision-making process given the customer does not require rigorous or moderate rational analysis of the purchase and he/she recures to their memory of personal feelings of previous experiences. In this context, AI might model the perception of individuals when they purchase products or services given they provide a review or feedback of their consumption experience using the most remarkable aspects [13,19,20].

In the digital age, it is essential to recognize the irreplaceable value of human contact, which is the main limitation of AI implementation in economic sectors. It cannot replicate the depth, empathy, understanding, and emotional connection that human touch provides [21]. AI becomes smart by understanding and replicating human movements, responding, and providing solutions. However, AI algorithms cannot give a gentle hug to the customer, a comforting pat on the back, hold someone’s hand during a difficult moment, or create a meaningful relationship with the client. AI excels at data collection, processing, and analysis, but fails to understand the nuances of nonverbal communication and the complexities of human interaction. AI-hedonic customer experience could capture the perception of clients. However, it cannot fully capture the core of human connection through human touch, which is the universal language that communicates emotions, fosters connections, and promotes overall well-being [22,23].

2.2.2. AI-Recognition Customer Service

Customer recognition shows respect and gratitude to individuals or other businesses who buy products or services of a firm. It is a powerful tactic in business strategy that will help to maintain a loyal and engaged customer base. The recognition aspect mentions qualities such as safety, importance, relation, being welcome, and a sense of beauty [19,24]. Nowadays, products and services include innovation factors, which are transformed into competitive advantages in the marketplace. These information technology advances, which also include AI, are linked to customer recognition given brands track customers’ purchase history, which identifies repeated and new customers according to their preferences and consumption. Therefore, the trend of customer identification intensifies the contest between brands, and firms show innovation in the design of products and services [25].

2.3. Customer Perception Factors and AI-Enabled Customer Experience

Customer perception factor compromises five factors of convenience in use (CON), personalization (PER), trust (TRU), customer loyalty (LOY), and customer satisfaction (SAT). AI-enabled customer experience compromises two factors: AI-hedonic customer experience (HCE) and AI-recognition customer service (RCS) [8,26,27,28,29,30].

2.3.1. Convenience in Use

Convenience is defined as the aptitude to achieve a goal in the limited time with the minimum resources. Convenience also refers to the savings in time and effort perceived by customers when purchasing and using services [31]. Customers show engagement in products and services when they perceive convenience in the use of the purchased good or service. Moreover, a convenient service or product is considered a solution to a need, which saves time and effort with the satisfaction of clients, who also are interested in a continuous relationship with the seller. Specifically, the convenience in the use of AI-enabled services will be categorized in three dimensions [32,33]: (i) availability of services, which includes 24/7 support and facility to contact the service/self-service, (ii) assistance and admission to real-time information, and (iii) proactive discussions between AI-powered bots and clients providing related information and assistance to the customer. All these AI-enabled aspects might improve time resolution, customer satisfaction, and engagement with the brand experience. Moreover, convenience in use increases the sustainable competitive advantage, especially in services where the core offerings are considered undifferentiated (e.g., services provided by banks) [34,35].

Convenience in use increases trust in the product, service, brand, and firm. Therefore, the perception of convenience in use influences positively the perception of customers when they assess the service utility, which also scores high given the reduced gap between product expectation and real product. Convenience in use reduces the customers’ sacrifice [36] given AI-enabled services can be used anytime and anywhere by customers. Furthermore, convenience in use is linked with consumers and their well-being with minimal effort, which also overcomes obstacles in customer’s consumption activities and increases customers’ comfort [37]. Therefore, following is the first set of hypotheses:

H1: 

Convenience in use will have a positive effect on the AI-enabled customer experience.

H1a: 

Convenience in use will have a positive effect on the AI-hedonic customer experience.

H1b: 

Convenience in use will have a positive effect on the AI-recognition customer experience.

2.3.2. Personalization

It is defined as the quality of personalized information that covers needs and preferences of a single customer producing a positive experience in individual consumption [35]. Personalization is possible because of the existence of AI techniques such as visual analytics, data mining, automation, machine learning, and robotics, which provide patterns of customer consumption data. Therefore, personalization is associated with AI-enabled services given the optimization of resources through algorithms and prediction models to make decisions. Specifically, personalization of online services can be measured by [38] (i) the user interface, which is the flexibility and functionality of screen design and overall presentation, (ii) content, which remarks the personalization of individual user’s profile using its information, preferences, and characteristics to send specific product or services offerings with prices, and (iii) interaction process, which includes AI interactions to approach users. Personalization by AI generates adapted content through data analysis from real-time user decisions. Access to the right user’s data might not affect the customer’s privacy since algorithms are programmed from interface data. However, data protection regulation represents a challenge to AI, firms, users, and governments.

Previous studies suggested that a high level of personalization of AI-enabled services is associated with (i) an increase in the brand’s competence [39], (ii) an increase in customers’ perception of firm quality [40], (iii) a decrease in sensitivity of customer sacrifice, and (iv) increase in positive firms’ attributions, which reinforce the trust of clients in the brand and firm [41]. Therefore, following is our second set of hypotheses:

H2: 

Personalization will have a positive effect on the AI-enabled customer experience.

H2a: 

Personalization will have a positive effect on the AI-hedonic customer experience.

H2b: 

Personalization will have a positive effect on the AI-recognition customer experience.

2.3.3. Trust

Trust involves benevolence and credibility of the firm, its products, and its services [42]. It is based on the benefits and the customer´s belief that the promises made by a product or service provided by a firm are reliable. Based on trust, customers share their personal information given they associated this process with confidentiality, which grows and develops the customer–firm relationship [43]. In the e-commerce sector, trust includes brand belief and the firm’s technology. Especially, firms need to consider the sensibility of the treatment of customer data in digital experiences [44]. Previous studies showed that AI involves trust. It ensures product or service acceptance, increases productivity by continuous progress, increases innovation in a firm, and develops technology [45]. Moreover, the novel dimensions of trust in AI-enabled customer service includes technology and its comprehension, brand, and purpose (confidence in intentions) [46,47]. Therefore, trust in AI is based on the transparency of functional logic, algorithm, and code (syntaxis).

Trust also includes the communication process of innovative technologies in the firms’ processes. This communication needs to be proactive to influence positively social acceptance of new technologies. The long-term relationship between the brand and customers depends on customers’ experience and trust, which shows a positive relationship in the current and subsequent experiences [16,48]. Therefore, following is our third set of hypotheses:

H3: 

Trust will have a positive effect on the AI-enabled customer experience.

H3a: 

Trust will have a positive effect on the AI-hedonic customer experience.

H3b: 

Trust will have a positive effect on the AI-recognition customer experience.

2.3.4. Customer Loyalty

Loyalty involves the steps of recurrently buying products and services and recommending them to others, which helps to grow business with word-of-mouth and referral marketing. It includes the occurrence and emotions of buying from a specific retailer and the previous experiences and attitudes of this process [49]. Specifically, AI-enabled customer experience is linked with customer loyalty given it prevents fraud by technology automatization and cyber security, determines customer requirements and behavior for better targeting, and discovers and manages industry data with machine learning applications, provoking the increase in firms’ competitive advantage [50]. Previous studies showed that customer loyalty might be empowered by AI techniques related to perceived value, cognitive trust, affective trust, and satisfaction of clients based on the service quality in the information systems, which enhance the AI-enabled customer experience [51,52,53]. Therefore, following is our fourth set of hypotheses:

H4: 

Customer loyalty will have a positive effect on the AI-enabled customer experience.

H4a: 

Customer loyalty will have a positive effect on the AI-hedonic customer experience.

H4b: 

Customer loyalty will have a positive effect on the AI-recognition customer experience.

2.3.5. Customer Satisfaction

Customer satisfaction is calculated by the difference between performance/realization/inequity and expectation/confirmation of the product or service [54]. It evaluates the behavior of clients after the purchase and quantifies the satisfaction level of the customer, which involves the service environment [55]. Firms that introduce AI in their processes might deliver the proactive and personalized services that customers want. For instance, applications and new interfaces are adopted by banking institutions to provide customer service, both include social and easy payment systems that integrate technology with legacy systems based on governance structure mechanisms. Therefore, there is a virtuous circle using AI-enabled customer experience given it can increase the significant value of a business by providing better service. Customer engagement might be exploited using AI techniques, which increases upsell and cross-sell opportunities with the reduction in cost-to-serve. Prior studies showed a positive relationship between customer satisfaction and AI-enabled customer experience, which also impacts positively the firm performance [56,57]. Therefore, following is our fifth set of hypotheses:

H5: 

Customer satisfaction will have a positive effect on the AI-enabled customer experience.

H5a: 

Customer satisfaction will have a positive effect on the AI-hedonic customer experience.

H5b: 

Customer satisfaction will have a positive effect on the AI-recognition customer experience.

To sum up, AI-enabled customer experience will be influenced by convenience in use, personalization, trust, customer loyalty, and customer satisfaction, given customers develop an impression about a product or service through advertisements, promotions, reviews, social media feedback, etc. All these appreciations might be entitled by the introduction of AI mechanisms in all firm’s processes given it increases the quality of service using customers’ perceptions, likes, and expectations. Moreover, AI might translate the customers’ feelings about products, services, and brands into their purchase intention and their willingness to pay. This valuable information allows managers to recognize the firm’s opportunities and challenges to improve all processes in a firm with competitive financial ratios. Therefore, our sixth set of hypotheses refers to the multiple effects of all determinants on AI-enabled customer experience:

H6: 

Customer perception factors will have a positive effect on the AI-enabled customer experience.

H6a: 

Customer perception factors will have a positive effect on the AI-hedonic customer experience.

H6b: 

Customer perception factors will have a positive effect on the AI-recognition customer experience.

2.3.6. Research Model

The research model shows the relationship between customer perception factors and the AI-enabled customer experience factors, including hedonic and recognition attributes in the Ecuadorian baking industry (Figure 1).

3. Methodology

3.1. Instrument and Data

A survey instrument was designed based on the constructs of the research. Respondents were asked to indicate their perception of the presence of AI in their virtual/mobile banking app. The questionnaire was designed using a five-point Likert scale and included typical demographic information. The instrument was pre-tested by researchers and customers (not included in the final sample). At the end of the pre-test, the questionnaire was modified to improve clarity. A series of meetings with experts was followed to measure the content validity and reliability of the instrument. Participants in this study were users of the Ecuadorian banking industry.

3.2. Procedure

Procedure of the survey was conducted with a sample size of 248 individuals using a convenient sampling method. We conducted a prior test to improve the reliability and validity of the measurement tool. Data from the survey were collected from November 2022 to December 2022. All invalid data were excluded, and 226 records were utilized for this study.

3.3. Measurement of Constructs

The effect of customer perception factors on the AI-enabled customer experience in the Ecuadorian baking industry was collected using an online survey. The study contains five constructs for customer perception (convenience in use, personalization, trust, customer loyalty, and customer satisfaction), and two constructs for AI-enabled customer experience (AI-hedonic customer experience and AI-recognition customer experience). The study employed multiple items to measure all constructs using five-point Likert scale for each item, displaying 1 and 5 for strongly disagree and strongly agree, respectively.

Table 1. Scale items of constructs.

Constructs	Items	Label	Related Literature
Socio-demographic information	Gender, age, marital status, level of education, occupation, monthly income, banking entity.	Nominal scale
AI-enabled customer experience (AIK)	Are you familiarized with the definition of AI? Do you think your bank uses AI to design the financial products and services?	AIK1, AIK2	[8,26]
Convenience in use (CON)	I can use the services of the virtual/mobile app of my bank whenever I want.	CON1	[27,28]
	The virtual/mobile app of my bank is available continuously and permanently.	CON2
	I value the possibility of using the virtual/mobile app of my bank from the comfort of my home.	CON3
	I am disturbed about the loss of control when I use the virtual/mobile app of my bank.	CON4
	I am concerned about the loss of confidentiality when I use the virtual/mobile app of my bank.	CON5
Personalization (PER)	The virtual/mobile app of my bank is personalized.	PER1	[26,29,30]
	The financial products and services of my bank are designed in concordance with my consumption preferences.	PER2
	The virtual/mobile app of my bank detects human errors in financial operations.	PER3
	Most of the financial processes in my bank are automated.	PER4
	The virtual/mobile app of my bank meets my expectations.	PER5
Trust (TRU)	The virtual/mobile app of my bank is reliable.	TRU1	[26,58,59]
	I am pleased with the electronic security of my bank.	TRU2
	The virtual/mobile app of my bank is secure.	TRU3
	The staff of my bank help me to solve problems with confidence.	TRU4
	The virtual/mobile app of my bank has good technical features.	TRU5
Customer loyalty (LOY)	I have an emotional link with my bank.	LOY1	[60,61,62,63]
	I feel recognized with my bank.	LOY2
	It is very difficult to change my bank.	LOY3
	I mention attributes of my bank.	LOY4
	I am a faithful client of my bank.	LOY5
Customer satisfaction (SAT)	Overall, I am content with the quality of service of my bank.	SAT1	[62,63]
	My bank is positioned at least in the third position in the national financial system.	SAT2
	Kindness and client’s attention are qualities of the staff of my bank.	SAT3
	When the staff of my bank answers my questions, they show knowledge and experience.	SAT4
	My bank offers individualized and personalized attention.	SAT5
AI-hedonic customer experience (HCE)	Using the virtual/mobile app of my bank is a memorable experience	HCE1	[26,50,64]
	The use of the virtual/mobile app of my bank is entertaining.	HCE2
	The use of the virtual/mobile app of my bank is exciting.	HCE3
	I feel comfortable using the virtual/mobile app of my bank.	HCE4
	I increase my learning activity when I use the virtual/mobile app of my bank.	HCE5
AI-recognition customer service (RCS)	The virtual/mobile app of my bank is the most important tool for all financial services.	RCS1	[26,50,65]
	The virtual/mobile app of my bank is designed in a respectful way toward the customer.	RCS2
	A welcome message is displayed when I use the virtual/mobile app of my bank.	RCS3
	The virtual/mobile app of my bank is secure.	RCS4
	The virtual/mobile app of my bank is well aesthetic and designed.	RCS5

presents the list of items for each construct and the previous studies that support each item and construct.

Google Forms and IBM SPSS Statistics 26 were the tools used to collect and process all the responses of the questionaries. A total of 248 questionnaires were recorded. The final sample consists of 226 records, given the duplicated observations and invalid responses.

This study is included in Section 4 and has the following subdivisions: demographic analysis, exploratory factor analysis (EFA), reliability analysis, validity analysis, correlation analysis, and regression analysis to prove the hypothesis of the study. Moreover, it employs thresholds for each test to validate the consistency of the sample and the measurements for each construct.

4. Empirical Results

4.1. Demographic Analysis

Table 2. Socio-demographic characteristics.

Variable	Classification	Percentage
Gender	Women	60.8%
	Men	39.2%
Age	26–35 years old	41.5%
	36–45 years old	31.5%
	18–25 years old	10.8%
	46–55 years old	8.5%
	Higher than 56 years old	7.7%
Marital status	Married	43.8%
	Single	38.5%
	Free union	10.0%
	Divorced or separated	6.9%
	Widower	0.8%
Academic trajectory	Masters and doctorate degrees	49.2%
	Junior college graduates	31.5%
	College graduates	17.7%
	Primary education	1.5%
Occupation	Private employees	56.9%
	Public employees	25.4%
	Own job and entrepreneur	12.3%
	Housewives and students	5.4%
Monthly income	USD 450.00–USD 1200.00	50.8%
	More than USD 2000.00	24.6%
	USD 1200.01–USD 2000.00	20.8%
	Lower than 450.00.	3.8%
Baking entity	Ecuadorian private bank	91.5%
	Ecuadorian public bank	8.5%

summarizes the socio-demographic characteristics of the sample. It is important to mention that 65.4 percent of respondents specified that they knew the definition of AI while 34.6 percent had never heard of the concept of AI. Moreover, 43.8% of the total respondents believe that their bank uses AI in the design of financial products and services while the remaining respondents (56.2%) were not completely sure of their bank using AI.

4.2. Exploratory Factor Analysis, Reliability, and Validity

The Cronbach’s alpha is used to assess the reliability of each scale. Alpha values over 0.7 indicate that all scales can be considered reliable [66]. For each of the item scales, factor analysis is used to reduce the total number of items to manageable factors. Principal components analysis is used to extract factors with eigenvalue greater than 1. Oblimin is used to facilitate interpretation of the factor matrix. Sampling adequacy measurement tests are also examined using Kaiser–Meyer–Olkin statistics to validate the use of factor analysis.

The study employed principal component analysis with 0.5 as the factor loading value. The omitted and removed items were CON1, CON2, CON3, PER3, PER4, TRU1, TRU2, TRU3, SAT1, SAT2, HCE1, HCE3, HCE5, and RCS4. These variables indicated segregated validity and lower internal consistency. The 60% of items were conserved in the analysis (initial number of items: 35 and final number of items: 21). EFA was conducted to confirm the validity of the measurement tool composition. A total of five sub-dimensions were set as customer perception factors such as “convenience in use” (two items), “personalization” (three items), “trust” (two items), “customer loyalty” (five items), and “customer satisfaction”(three items) while two sub-dimensions were employed as AI-enabled customer experience such as “AI-hedonic customer experience” (two items) and “AI-recognition customer service” (four items). Therefore, a total of 21 items were analyzed.

Table 3. Exploratory factor analysis results and reliability analysis.

Constructs	Label	Factor Loadings	AVE	Composite Reliability	Cronbach’s Alpha	Std. Deviation	Variance
CON	CON4	0.915	0.817	0.899	0.782	1.295	1.676
	CON5	0.892				1.339	1.794
PER	PER1	0.717	0.519	0.759	0.736	0.641	0.411
	PER2	0.844				0.959	0.920
	PER5	0.573				0.840	0.705
TRU	TRU4	0.790	0.587	0.740	0.630	1.005	1.009
	TRU5	0.742				0.721	0.520
LOY	LOY1	0.784	0.562	0.865	0.897	1.257	1.579
	LOY2	0.757				1.170	1.368
	LOY3	0.710				1.189	1.415
	LOY4	0.745				1.167	1.362
	LOY5	0.750				1.159	1.344
SAT	SAT3	0.790	0.583	0.806	0.881	1.062	1.128
	SAT4	0.813				0.947	0.896
	SAT5	0.681				1.067	1.139
HCE	HCE2	0.805	0.608	0.756	0.659	1.124	1.264
	HCE4	0.754				1.013	1.027
RCS	RCS1	0.698	0.521	0.812	0.780	1.001	1.002
	RCS2	0.799				0.878	0.770
	RCS3	0.702				1.035	1.070
	RCS5	0.681				0.843	0.710
Kaiser–Meyer–Olkim (KMO) = 0.894 Significance of Bartlett’s Test of Sphericity = 0.000 Extraction Sums of Squared Loadings (Cumulative Variance %) = 71.053%.

Note: N = 226. Extraction method: principal component analysis. Rotation method: Oblimin. Factor extraction criteria: eigenvalue (1, 0). AVE = average variance extracted.

shows the EFA results and reliability analysis.

First, the correlation between variables was determined using the Kaiser–Meyer–Olkin (KMO) test and the Bartlett’s sphericity test. The significance of Kaiser–Meyer–Olkim (KMO) and the significance of Bartlett´s sphericity test were 0.894 and 0.000 (p < 0.05), respectively, indicating sampling adequacy given the KMO value is close to 1, which means that the correlation between variables is strong. Generally, 0.80 or higher is fairly adequate, 0.70 to 0.79 is adequate, 0.60 to 0.99 is normal, 0.5 to 0.59 is undesirable, and 0.5 or less is unacceptable. And Bartlett’s test judged that each variable was independent to a certain degree and suitable for factor analysis if the p value was less than 0.05 when judging the correlation between variables. Finally, the consistency and validity of the questionnaire were analyzed using Oblimin factor rotation. Principal component analysis is a method of factor analysis and is a method of extracting factors with eigenvalues greater than 1. In general, if the factor loading is 0.4 or more, it can be judged that the variable is significant.

Cronbach’s alpha level was between 0.630 and 0.897. The suggested level for composite reliability is 0.7. Its value in the study ranged from 0.740 to 0.899. The recommended value for average variance extracted (AVE) is 0.5. Its level fluctuated from 0.519 to 0.817. Table 3 exhibits tolerable discriminant validity for all constructs.

4.3. Discriminant Validity Analysis

Table 4. Discriminant validity analysis of confirmatory factor analysis.

Var	Items	Mean	SD	Correlations
				CON	PER	TRU	LOY	SAT	HCE	RCS
CON	2	3.669	1.194	(0.904)
PER	3	4.272	0.666	0.291 ***	(0.720)
TRU	2	4.042	0.747	0.221 ***	0.620 ***	(0.766)
LOY	5	3.520	1.000	0.217 ***	0.459 ***	0.504 ***	(0.749)
SAT	3	3.846	0.923	0.323 ***	0.528 ***	0.694 ***	0.561 ***	(0.763)
HCE	2	3.577	0.924	0.310 ***	0.549 ***	0.568 ***	0.640 ***	0.591 ***	(0.780)
RCS	4	4.138	0.731	0.346 ***	0.557 ***	0.629 ***	0.426 ***	0.618 ***	0.618 ***	(0.722)

Note: SD = standard deviation. Values in parenthesis are square root of AVE. *** indicates significance at the 1% level.

shows the discriminant validity analysis of confirmatory factor analysis and correlation between constructs.

The compound score of PER was 4.272, which was the maximum value of client perception factors, ranging from 3.520 to 4.042. This result shows that the main component for Ecuadorian customers is the personalization provided by the banking industry. Personalization is presented in virtual/mobile banking apps according to the user’s consumption preferences, which also meets the user’s expectations.

The TRU composite score is 4.042, showing a high level of trust of customers in the bank. This finding is supported by the customer perception of confidence, interest, and certitude in the technical features shown in the virtual/mobile bank app. The respondents exhibited a concise sensibility for satisfaction, SAT component (µ = 3.846), showing that the staff of the bank demonstrates kindness (µ = 3.815) and knowledge (µ = 3.815) by providing individualized and personalized (µ = 3.908) attention to their customers. The composite score for CON was 3.669, suggesting that customers are concerned about the loss of privacy (µ = 3.546) and control (µ = 3.792) when they use their virtual/mobile bank app. The conglomerate score for LOY was 3.520. However, the loyalty component is the unique factor that preserves all items given its internal consistency and validity. These results reveal that customers believed to have an emotional link with their bank (µ = 3.308), also identified (µ = 3.492) and mentioned attributes (µ = 3.646) of their financial institution, expressed not to change their bank (µ = 3.500), and finally, considered them as loyal bank customers (µ = 3.654).

The factors of AI-enabled customer experience evidenced high composite scores, such as 3.577 for HCE and 4.138 for RCS. These findings suggest that customers perceive the virtual/mobile bank app as entertaining (µ = 3.262), comfortable to use (µ = 3.892), respectful towards clients (µ = 4.208), welcoming at the first step of use (µ = 4.085), and with a well-designed aesthetic (µ = 4.185). Therefore, customers considered the virtual/mobile bank app as the most important tool for all financial services (µ = 4.077). Moreover, there were no multicollinearity problems and the AVE’s square root for each construct was larger than the correlation coefficients between variables [67].

Next, a discriminant validity analysis was conducted after an intensive validity analysis. Discriminant validity was tested by considering AVE and Pearson’s correlation values. Table 4 shows the correlations among constructs. The table shows the square root of the AVE values on the diagonal. The square root of the AVE values exceeded 0.720. If the correlation coefficient exceeds its value, then the validity between each component of the concept is secured. The results showed that the square root of each variable’s AVE exceeded the correlation value for each variable, demonstrating sufficient discriminant validity. Thus, all five factors of CON, PER, TRU, LOY, and SAT were net nested into HCE. Moreover, the result of discriminant validity analysis showed that all correlation coefficients were significant at the significance probability (p) *** < 0.01 level.

4.4. Regression Analysis

Table 5. Regression results of hypothesis tests.

Variables	AIK						HCE						RCS
CON	0.149 ***					0.083 **	0.085 **					0.018 **	0.212 ***					0.158 ***
	(2.770)					(2.386)	(2.255)					(2.159)	(4.167)					(4.268)
	[1.945]					[3.023]	[1.421]					[2.732]	[4.682]					[3.126]
PER		0.686 ***				0.258 ***		0.761 ***				0.276 **		0.611 ***				0.239 ***
		(8.786)				(3.182)		(7.429)				(2.474)		(7.589)				(2.769)
		[3.333]				[2.558]		[2.135]				[5.614]		[1.727]				[1.205]
TRU			0.659 ***			0.203 **			0.702 ***			0.150 **			0.616 ***			0.256 ***
			(9.950)			(2.406)			(7.801)			(2.290)			(9.161)			(2.844)
			[1.980]			[5.177]			[1.823]			[1.181]			[2.658]			[1.260]
LOY				0.452 ***		0.177 ***				0.591 ***		0.353 ***				0.312 ***		0.002 **
				(8.612)		(3.451)				(9.419)		(4.975)				(5.335)		(2.040)
				[5.179]		[2.059]				[2.059]		[1.118]				[2.688]		[2.461]
SAT					0.053 ***	0.207 ***					0.592 ***	0.187 **					0.490 ***	0.228 ***
					(10.191)	(3.131)					(8.292)	(2.044)					(8.886)	(3.228)
					[2.732]	[2.688]					[5.614]	[2.380]					[1.900]	[3.169]
Constant	1.403 ***	0.926 ***	1.193 ***	2.268 ***	0.210 ***	0.817 **	3.890 ***	0.325 *	0.738 **	1.495 ***	1.299 ***	0.942 **	4.916 ***	1.527 ***	1.648 ***	3.041 ***	2.255 ***	1.777 ***
	(21.281)	(2.741)	(4.382)	(11.820)	(8.466)	(2.561)	(14.824)	(1.732)	(1.995)	(6.509)	(4.599)	(2.324)	(25.075)	(4.383)	(5.964)	(14.214)	(10.349)	(5.222)
Adj. R2	0.249	0.371	0.432	0.362	0.444	0.608	0.212	0.296	0.317	0.405	0.344	0.515	0.213	0.305	0.391	0.275	0.377	0.536
F	17.673 ***	77.190 ***	99.008 ***	74.166 ***	103.852 ***	41.017 ***	17.575 ***	55.185 ***	60.854 ***	88.722 ***	68.757 ***	28.382 ***	17.364 ***	57.591 ***	83.927 ***	28.458 ***	78.960 ***	30.858 ***

Note: Beta corresponds to unstandardized coefficients. Numbers inside the parenthesis are t-statistics. Numbers inside the brackets are the variance inflation factors. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

presents the results of individual linear regressions to test the effect of five customer perception factors (individual model) as well as customer perception construct (aggregated model) on AI-enabled customer experience in the Ecuadorian banking industry. The adjusted R-square ranged from 0.212 (H1a result) to 0.608 (H6 result). All hypotheses of the study are proved. Fundamentally, H6 is supported, suggesting that AI-enabled customer experience is positively affected by customer perception factors (β = 0.083, 0.258, 0.203, 0.177, and 0.207; p < 0.05, p < 0.01, p < 0.05, p < 0.01, and p < 0.01, respectively). These results show that the customers evaluate the convenience in use, personalization, trust, loyalty, and satisfaction in the design and use of virtual/mobile bank apps, which bolsters a long-term business and financial relationship between clients and banks [48]. The integration of AI in the products and services offered by banks improves the customer experience given the influence of innovation and digital technologies in the service process [26]. AI-enabled customer experience might be enhanced when customers perceive a secure and reliable environment, which also includes additional support provided by the financial institution. Moreover, AI-enabled customer experience involves human emotions and a feeling of acknowledgment, which are the main components of the hedonic and recognition customer experience. Therefore, customers have a positive attitude toward AI given their aptitude and skills to comprehend and use AI-based financial products and services [2,26,50].

All customer perception factors were influenced individually (at least at the 5% level) by AI-enabled, hedonic, and recognition customer experience. Therefore, all hypotheses from H1 to H5b are supported given the significant individual coefficients. Particularly, convenience in use impacts AI-enabled experience by the facility to conduct a self-service navigation in a virtual/mobile bank app, which is also complemented by the assistance offered to the customers through bank staff and AI-powered bots. The findings of this study are aligned with previous research. The use of powerful AI mechanisms increases the service utility since users have access to real-time information and support, which also improves time problem resolution and increases brand engagement [32,36]. Moreover, personalization as a customer perception factor is associated with AI-enabled customer experience given financial products and services are designed using real-time personal and behavioral user information to meet the preferences and consumption decisions of bank clients [38,40].

Running individual regressions on trust, customer loyalty, and customer satisfaction, each of them exhibited a significant positive relationship (at the 1% level) with AI, HCE, and RCS components. Bank clients mentioned that their banks own attractive and modern amenities, which include high level of technology, innovation, and security, concluding that customers are influenced by and identified with attitude and performance of their financial institutions [45].

The results of this study are associated with preceding literature. Authors mentioned that AI-enabled customer experience and service might be influenced by personalization, technology, transparency, cognitive trust, affective trust, confidence, purpose, satisfaction, and service quality in the information systems [46,47,51,52,53]. Also,

Table 6. Multiple regression results of hierarchical relationship test.

Variables	HCE						RCS
CUP	1.066 ***						0.652 ***
	(9.734)						(6.606)
CON		0.085 **	0.047 *	0.029 **	0.011 **	0.018 **		0.212 ***	0.182 ***	0.165 ***	0.162 ***	0.158 ***
		(2.255)	(1.820)	(2.527)	(2.225)	(2.159)		(4.167)	(4.304)	(4.296)	(4.223)	(4.268)
PER			0.754 ***	0.442 ***	0.301 ***	0.276 **			0.582 ***	0.290 ***	0.269 ***	0.239 ***
			(7.314)	(3.584)	(2.677)	(2.474)			(7.668)	(3.331)	(3.025)	(2.769)
TRU				0.452 ***	0.265 **	0.150 **				0.424 ***	0.397 ***	0.256 ***
				(4.093)	(2.567)	(2.290)				(5.441)	(4.846)	(2.844)
LOY					0.398 ***	0.353 ***					0.058 **	0.002 **
					(5.838)	(4.975)					(2.066)	(2.040)
SAT						0.187 **						0.228 ***
						(2.044)						(3.228)
Constant	0.550 **	0.890 *	0.530 **	0.435 **	0.643 **	0.942 **	1.614 ***	4.916 ***	2.322 ***	1.793 ***	1.777 ***	1.777 ***
	(2.283)	(1.824)	(2.041)	(2.069)	(2.321)	(2.324)	(4.180)	(5.075)	(6.187)	(5.081)	(5.035)	(5.222)
R2	0.225	0.012	0.305	0.387	0.518	0.534	0.254	0.119	0.398	0.513	0.517	0.554
Var. R2		0.012	0.293	0.082	0.131	0.016		0.119	0.279	0.115	0.004	0.037
Adj. R2	0.21	0.004	0.294	0.372	0.503	0.515	0.248	0.113	0.389	0.501	0.502	0.536
F	94.744 ***	1.575 ***	27.859 ***	26.462 ***	33.579 ***	28.382 ***	43.635 ***	17.364 ***	41.998 ***	44.170 ***	33.448 ***	30.858 ***
Var. F		1.575	53.496	16.755	34.086	4.180		17.364	58.792	29.600	1.137	10.417

Note: CUP represents the sum of the five customer perception factors. Beta corresponds to unstandardized coefficients. Numbers inside the parenthesis are t-statistics. Var R² is an incremental ratio of R² value change. Var. F is an incremental ratio of F value change. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

shows whether all five factors of the costumer perception is nested or hierarchical. We showed that when we introduced each customer perception factor one-by-one, the value of R² increased, generating the highest values of 0.534 and 0.554 for HCE and RCS models, respectively.

All these characteristics create a competitive advantage for institutions and create loyal customers given their continuous AI service satisfaction provoked by anticipated customers’ perceptions and expectations.

Table 7. Summary of study hypotheses test results.

Hypothesis Statements		Decisions
H1: Convenience in use will have a positive effect on the AI-enabled customer experience.		Fully supported (at least at the 1% level)
	H1a: Convenience in use will have a positive effect on the AI-hedonic customer experience.	Supported (at least at the 5% level)
	H1b: Convenience in use will have a positive effect on the AI-recognition customer experience.	Fully supported (at least at the 1% level)
H2: Personalization will have a positive effect on the AI-enabled customer experience.		Fully supported (at least at the 1% level)
	H2a: Personalization will have a positive effect on the AI-hedonic customer experience.	Fully supported (at least at the 1% level)
	H2b: Personalization will have a positive effect on the AI-recognition customer experience.	Fully supported (at least at the 1% level)
H3: Trust will have a positive effect on the AI-enabled customer experience.		Fully supported (at least at the 1% level)
	H3a: Trust will have a positive effect on the AI-hedonic customer experience.	Fully supported (at least at the 1% level)
	H3b: Trust will have a positive effect on the AI-recognition customer experience.	Fully supported (at least at the 1% level)
H4: Customer loyalty will have a positive effect on the AI-enabled customer experience.		Fully supported (at least at the 1% level)
	H4a: Customer loyalty will have a positive effect on the AI-hedonic customer experience.	Fully supported (at least at the 1% level)
	H4b: Customer loyalty will have a positive effect on the AI-recognition customer experience.	Fully supported (at least at the 1% level)
H5: Customer satisfaction will have a positive effect on the AI-enabled customer experience.		Fully supported (at least at the 1% level)
	H5a: Customer satisfaction will have a positive effect on the AI-hedonic customer experience.	Fully supported (at least at the 1% level)
	H5b: Customer satisfaction will have a positive effect on the AI-recognition customer experience.	Fully supported (at least at the 1% level)
H6: Customer perception factors will have a positive effect on the AI-enabled customer experience.		Supported (at least at the 5% level)
	H6a: Customer perception factors will have a positive effect on the AI-hedonic customer experience.	Supported (at least at the 5% level)
	H6b: Customer perception factors will have a positive effect on the AI-recognition customer experience.	Supported (at least at the 5% level)

shows a summary of all hypotheses test results.

5. Discussion

Banking industry has included advanced AI technologies in all their financial operations (evolving to Backing 4.0). Their adoption comprises lower operating costs and acceleration of their inclusion in banking services. This study aims to analyze the factors of customer perception that influence AI-enabled customer experience in the Ecuadorian banking environment. The results showed that the strength of customer perception factors was associated with improved AI-enabled customer experience in a bank, which includes financial technologies with AI algorithms in their products and services [2].

A clear example is the design and incorporation of AI in bank’s virtual/mobile apps to provide clients with options to make transactions, check their balances, and access financial suggestions. In this case, customer perception of the virtual/mobile app in financial institutions measures factors such as convenience in use, personalization, trust, customer loyalty, and customer satisfaction. These factors are key builders of the AI-enabled customer experience of the institution given clients score a financial product or service in terms of their expectations, needs, and facilities.

In light of the theory, this study verified that customer perception has an influence on AI-enabled customer experience, specifically on AI-hedonic customer experience and AI-recognition customer experience in the Ecuadorian Banking environments. These results support the arguments that convenience in use, personalization, trust, loyalty, and satisfaction are nested into the customer perception construct in these research settings. This study findings are important since it is possible to address that more customer perception in Ecuadorian Banking companies have a better positive impact on AI-enabled experienced customers. From the study, higher customer perception and awareness may increase a better AI-assisted and AI-enabled experience. It can be argued that the AI-enabled experience in Ecuadorian Banking environments that fulfill the needs of their customer perception regarding convenience in use, personalization, trust, loyalty, and satisfaction are positively affected, and they have a competitive advantage compared to their competitors.

The implementation method of our findings in the banking industry will depend on the information and experience of the client. This study provides evidence of the transcendental factors perceived by users in a virtual/mobile app in a financial institution. On the other hand, banks need to understand customers’ preferences and needs by analyzing clients’ information to design personalized customer service and support users’ queries in real-time. Therefore, the banking industry needs to assess the security and privacy of customer data, feed AI training models to incorporate relevant variables, and design an easy-to-use chat interface or virtual platform to provide an efficient and convenient user experience in banking and financial services. Moreover, even when similar studies have been developed in other countries in Asia, Europe, or North America, the results obtained could not be directly extended to the Ecuadorian use case. With our study, we have proved that a positive relationship, similar to the one found in other countries, exists in the Ecuadorian case. In addition, our study provides new insights into the banking sector that can be used to create competitive advantages when investing in and implementing AI in bank services, knowing beforehand that these services will be welcomed by clients as their user experience will improve.

AI-enabled customer experience might retain and increase loyal customers while AI changes the dynamics of how the world of business is run. Therefore, the banking industry should always be searching for ways to improve the customer experience by using AI in its processes. There are advantages and disadvantages of AI in customer experience. On one hand, the following are the pros: (i) personalized content and messages that display specific content according to the user’s profile by using AI algorithms, which crunch data on each customer and deliver tailored content based on the data, (ii) reduced costs by the diminution of workers and the implementation of chatbots that interact with customers in real-time and provide immediate responses, and (iii) advantage over competitors through a personalized, streamlined, and data-driven customer experience. On the other hand, the following are the cons of AI-enabled customer experience: (i) loss of human touch as customers want to interact with humans instead of a robot, (ii) wrong data to make decisions when data cleaning is not implemented as an important process in the firm, and (iii) cyberattacks and cybersecurity vulnerability.

This study involved the Ecuadorian banking sector, which is composed of 24 private banks. The financial services activities represented 3.6% of the Ecuadorian Real Gross Domestic Product (GDP). Through this study, the following means of design and implementation along with practical ways of applying AI in real settings are suggested.

Firstly, in the case of Ecuadorian banks, which is still in its introductory stage, the top management’s understanding and involvement in AI is of utmost importance. To this end, it is necessary to specify the introduction and implementation plan through a bank-wide comprehensive plan. The bank’s strategic planning department should draw up an action plan and design a detailed implementation plan for each department in a bank.

Second, business guidelines for technology and information exchange between branches must be adopted. There will be many difficulties in the job of system construction and mutual interface. In order to solve this problem, selecting a system suitable for the bank’s business scale and business size and focusing on the characteristics of the use of the system are closed related expertise and information collection. Thus, a systematic and organized manner should be carried out. Furthermore, full awareness of the function and role of the system and the training and deployment of experts with appropriate abilities are essential. For a bank, continuous education and training must be provided by specialist system vendors.

Finally, AI helps banks to personalize products and services, which also engage customers given the deep analysis of the client’s preferences databases with high-speed data processing. Mathematical complex algorithms are employed in the design and implementation of digital financial services, showing efficiency and rapidity compared to the traditional banking methods. Therefore, the application of AI technologies helped banks in the following ways: (i) to prevent fraud, (ii) to increase their reliability and accuracy, (iii) to show a direct impact on their financial operations, (iv) to accelerate automation, blockchain, and Fintech, (v) for data analysis, (vi) in asset management and risk management, and (vii) in customer service.

Through this study, the results revealed that five customer perception factors (convenience in use, personalization, trust, customer loyalty, and customer satisfaction) affect positively and significantly (at least at the 5% level) AI-enabled customer experience, hedonic customer experience, and recognition of customer service. The findings are consistent with [2,26,52,68], which revealed that AI-based financial services depend on the digital facilities provided by financial institutions and the customer impression of tangible and intangible banking services with AI. Therefore, the six sets of hypotheses were corroborated in our study given the integration of AI in financial services might cause a shift in how customers perceive their experiences.

AI-enabled customer experience is built by an integrated understanding of the customer perspective implementing innovative technologies and strategies. This process generates a customer-centric view of AI showing how the AI capabilities are experienced by customers. Therefore, customers may increase their service quality standards of products and services, which may also increase the market competitiveness, generating a permanent growth in the expectations and demands of clients, which may be improved by the implementation of AI mechanisms.

Finally, the method chosen to collect the data needed in this study, online surveys, can generate some bias in the results. The specific characteristics of the target population such as the age, gender, education, among others can be under or overrepresented. This can happen even when the sampling plan is well designed, as this method depends on the willingness and interest of the participants to respond to the survey. A special concern is the underrepresentation of population groups who lack the means to respond to online surveys, for example, lack of internet connection or familiarity with the topic. To overcome this last problem, the questions in the survey used in this study have been formulated in a simple manner to avoid misunderstanding. Moreover, no similar studies have been developed for the Ecuadorian case, so contrasting results are not possible, but the findings of this study show consistency with prior literature in other countries, which can be used to validate the obtained results.

6. Conclusions

We analyze the factors of customer perception that influence AI-enabled customer experience in the Ecuadorian banking environment. Using 226 records from a self-designed online questionnaire, the results revealed that AI-enabled customer experience, AI-hedonic customer experience, and AI-recognition customer service are affected positively by customer perception factors (individual and joint effect, e.g., convenience in use, personalization, trust, customer loyalty, and customer satisfaction) at least at the 5% level of significance. The positive relationship between customer perception factors and AI-enabled customer experience is supported by preceding studies‘ results in Asian countries [2], European cities [26], India [27], the United Kingdom, Canada, Nigeria, and Vietnam [10]. Furthermore, customer perception factors may be improved by AI techniques since most financial services incorporate AI algorithms and may identify the type of customer by recollecting, processing, and analyzing the purchases based on customer behavior.

Convenience in use, personalization, trust, customer loyalty, and customer satisfaction are the palpable and intangible factors that impact customer perception, while AI-enabled customer experience is formed by AI-hedonic customer experience and AI-recognition customer service. All factors empowered the AI services provided by financial institutions. Banks offer undifferentiated products and services, and thus, customer perception factors are the unique differentiated elements between entities, creating a major competitive advantage, which also positively motivates the development of digital banking products and services.

This paper provides a complete and significant statistical and econometrical model of factors that compose customer perception and it measures its positive influence on AI-enabled customer experience. Second, this study showed theoretical and practical evidence of the relationship between customer perception factors and AI-enabled customer experience in the Ecuadorian banking industry, which has not been deeply analyzed in the Ecuadorian context. Third, this study showed the importance of introducing AI in the design of financial products and services using as fundamental input the customer information about purchase behavior and preferences of consumers.

This study highlights the possible implications related to AI implementation in economic sectors. Automated systems will make some jobs obsolete, generate unemployment, and modify the viability of social security systems [44,69]. Therefore, it is crucial to adapt educational and training systems to the new digital skills requirements, while the current and future workforce needs to re-skill and up-skill themselves. Moreover, AI also has implications on cybersecurity, where applications play an important role in defensive and offensive cyber measures. AI works with large amounts of data, while the main concern is its privacy and protection. Thus, AI virtual/mobile apps must ensure data integrity, privacy, and confidentiality. Furthermore, AI systems involve judgments and decision-making processes that could replace the human process. However, there is a need to combine ethics training for AI practitioners, which might be consistent with existing laws, social norms, and ethics. Finally, regulatory approaches to AI should be put in place to identify principles for the development and application of AI, provide advice to the government, and encourage public dialogue.

The main limitation of the study is the inexistence of a customer perception index for the Ecuadorian economic sectors to compare and supplement our findings. Moreover, the study focused principally on virtual/mobile bank apps and digital financial services that incorporate AI in their design and execution. However, AI technologies involve also other financial products and services. The study only emphasized the most demanded AI financial services developed by Ecuadorian banks for their customers, to compare similar banking services between different institutions. Finally, in the Ecuadorian context, most customers (56.2% of respondents) did not differentiate between the presence or absence of AI in banking. Therefore, there is ignorance about the power of AI for developing a business. For future research, the authors recommend performing a longitudinal analysis using quantitative data to measure the impact of AI-enabled customer experience on the financial performance of Ecuadorian banks. Moreover, the quality evaluation of the used algorithms in virtual/mobile bank apps can be part of future work to understand the degree of development of AI in the banking services in Ecuador.