Artificial intelligence, customer-centric transformation, and autonomous finance

Abstract

Artificial intelligence is fundamentally reshaping the structure, logic, and competitive dynamics of the banking industry. This paper develops a conceptual framework—the Intelligence–Ecosystems–Autonomy (IEA) framework—to explain how AI-driven transformation reconfiguring banking as an interconnected system of cognitive, networked, and autonomous capabilities is. Rather than treating digital transformation as a linear progression of technological adoption, the framework conceptualizes it as a co-evolutionary process in which intelligence, ecosystem integration, and decision autonomy mutually reinforce one another.

The Intelligence dimension captures the emergence of AI-enabled cognitive financial systems that augment and increasingly participate in decision-making processes. The Ecosystems dimension describes the shift from firm-centric banking toward platform-based financial architectures characterized by interoperability, data sharing, and co-created value. The Autonomy dimension reflects the rise of agentic and algorithmic systems capable of executing financial decisions with limited human intervention, reshaping traditional governance and control structures.

By integrating these three dimensions, the paper argues that banking transformation is best understood as an architectural reconfiguration rather than incremental digital improvement. The framework highlights how advances in AI intelligence enhance ecosystem coordination, how ecosystems amplify machine learning capabilities through data expansion, and how both jointly enable increasing levels of system autonomy.

The paper contributes to digital transformation and financial services literature by offering a unified model that explains how AI reshapes not only operational efficiency but also the fundamental organization of financial systems, decision rights, and competitive advantage. It further identifies emerging tensions related to governance, accountability, interpretability, and systemic risk in increasingly autonomous financial environments.

1. Introduction

The global banking industry is undergoing a profound transformation driven by the convergence of digital technologies, artificial intelligence (AI), and evolving customer expectations. What was initially characterised as incremental digitalisation has developed into a systemic reconfiguration of business models, organisational processes, and competitive dynamics. Digital transformation in this context is not merely the adoption of new technologies but a fundamental shift in how financial institutions create, deliver, and capture value across organisational and institutional boundaries (Vial, 2019). In banking, these developments are particularly consequential, as technological innovation is reshaping both customer interfaces and core operational infrastructures, thereby redefining strategic positioning (Gomber et al., 2018; Boot et al., 2021).

Artificial intelligence has emerged as a central driver of this transformation. AI technologies enable real-time data processing, predictive analytics, and adaptive decision-making, significantly enhancing operational efficiency and organisational performance (Brynjolfsson and McAfee, 2017; Brynjolfsson et al., 2021). In banking, AI is increasingly embedded in core functions such as fraud detection, credit risk assessment, and customer service, where it improves accuracy, speed, and scalability (Huang and Rust, 2021; Rai, 2020). At the customer interface, AI facilitates personalised interactions, conversational banking, and predictive service delivery, thereby strengthening customer engagement and relationship depth (van Doorn et al., 2017; Wedel and Kannan, 2016). More fundamentally, AI is shifting banking from rule-based automation toward adaptive, learning systems capable of supporting complex and context-sensitive decision processes—an evolution often described as augmented intelligence (Raisch and Krakowski, 2021).

Despite these advances, the academic literature remains fragmented. Existing research has largely focused on isolated dimensions of transformation, including digital maturity, fintech innovation, and AI-enabled service automation. Studies on digital maturity emphasise the role of technological infrastructure and data integration in enhancing firm performance (Nicoletti, 2017; Boot et al., 2021), while fintech research highlights the impact of new entrants and innovation dynamics on financial markets (Haddad and Hornuf, 2019). Similarly, research on AI in services has explored its effects on customer experience and operational efficiency (Huang and Rust, 2021). While these streams provide valuable insights, they insufficiently capture the interdependencies between technological capabilities, organisational transformation, and strategic positioning. As a result, there is a lack of integrative frameworks that explain how these dimensions jointly reshape the logic of banking.

This limitation is increasingly problematic in light of emerging developments that extend beyond conventional digitalisation. First, advances in generative AI and machine learning are enabling cognitive process automation and context-aware decision support, allowing organisations to move beyond deterministic workflows toward adaptive operational systems (Davenport and Ronanki, 2018; Jöhnk et al., 2021). Second, the rise of platform-based ecosystems and embedded finance is transforming banks from standalone service providers into orchestrators of interconnected value networks (Parker et al., 2016; Zetzsche et al., 2018). Third, the emergence of agentic and semi-autonomous systems—capable of initiating and executing financial transactions—signals a transition toward increasingly autonomous financial infrastructures (Gomber et al., 2018; Vial, 2019). Together, these developments indicate that banking is entering a new phase characterised by intelligence, ecosystem integration, and partial autonomy.

At the same time, broader structural forces reinforce this transformation. Demographic change, particularly population ageing, is increasing demand for complex financial services such as retirement planning, decumulation strategies, and intergenerational wealth transfer (Horneff et al., 2013; OECD, 2023). These shifts necessitate more integrated, lifecycle-oriented advisory models that move beyond product-centric approaches toward holistic financial well-being ecosystems (Brüggen et al., 2017). In parallel, competitive pressures from fintech firms and non-bank actors continue to intensify, accelerating innovation and challenging traditional banking models (Haddad and Hornuf, 2019). These dynamics highlight the need for a holistic understanding of how technological, organisational, and environmental factors interact in shaping the future of banking.

Against this backdrop, this paper develops an integrative conceptual framework to explain contemporary banking transformation. Specifically, it proposes the IEA framework, which captures the transition along three interrelated dimensions: Intelligence, Ecosystems, and Autonomy. First, the paper conceptualises the shift from digitalisation to intelligence, emphasising the role of AI in enabling data-driven, adaptive, and context-aware decision systems (Brynjolfsson and McAfee, 2017; Raisch and Krakowski, 2021). Second, it examines the transition from product-centric models to ecosystem-based service architectures, in which value is co-created across organisational boundaries (Vargo and Lusch, 2008; Parker et al., 2016). Third, it explores the emergence of autonomous financial systems, highlighting the implications of agentic technologies for decision-making, governance, and competition (Rai, 2020; Gomber et al., 2018).

By integrating these dimensions, the paper makes three contributions to the literature. First, it develops a structured conceptual framework that links technological capabilities with organisational transformation and strategic outcomes in banking. Second, it advances emerging concepts—such as agentic commerce and AI visibility—that remain underexplored in existing research. Third, it identifies key risks and governance challenges associated with AI-driven transformation, including issues of trust, explainability, algorithmic bias, and digital identity (Rai, 2020; Kietzmann et al., 2020; Chesney and Citron, 2019).

The remainder of the paper is structured as follows. Section 2 examines digital maturity as a foundational capability. Section 3 analyses demographic change and its implications for pension and wealth advisory. Section 4 explores the role of generative AI in banking operations, while Section 5 introduces agentic commerce and autonomous financial ecosystems. Section 6 discusses AI-enabled workplace transformation, followed by Section 7 on customer-centric operating models. Section 8 addresses emerging risks related to digital identity and deepfakes, and Section 9 examines AI visibility as a new dimension of competition. Section 10 develops the theoretical implications of these transformations through the IEA framework, and Section 11 concludes.

2. Digital Maturity as a Strategic Capability

Digital maturity has emerged as a central determinant of competitiveness in the banking sector. Rather than reflecting the mere adoption of digital channels, digital maturity captures an institution’s ability to integrate technological infrastructure, data capabilities, and organisational processes into a coherent and adaptive system of value creation (Vial, 2019). As such, digital maturity constitutes a strategic capability, enabling firms to sense, seize, and transform in response to environmental change, consistent with the dynamic capability’s perspective (Teece et al., 1997; Warner and Wäger, 2019).

Empirical evidence indicates that higher levels of digital maturity are associated with improved organisational performance, including enhanced operational efficiency, increased innovation capacity, and superior customer acquisition outcomes (Nicoletti, 2017; Boot et al., 2021). In banking, digitally mature institutions are better positioned to leverage data analytics, automate processes, and deliver seamless, integrated customer experiences across multiple channels (Gomber et al., 2018). However, these benefits are unevenly distributed. Many incumbent banks struggle to translate digital investments into sustained competitive advantage due to legacy systems, organisational inertia, and fragmented data architectures (Vial, 2019; Jöhnk et al., 2021). This distinction underscores that digital maturity is not equivalent to digital adoption; it requires systemic integration rather than isolated technological upgrades.

A critical dimension of digital maturity lies in the transition from front-end digitalisation to backend transformation. Early phases of digital banking focused primarily on customer-facing innovations, such as mobile applications and online platforms. While these developments improved accessibility and convenience, they did not fundamentally alter the underlying logic of banking operations. More recent research emphasises that competitive advantage increasingly depends on backend capabilities, including cloud computing, application programming interfaces (APIs), and integrated data infrastructures (Alt et al., 2018; Gomber et al., 2018). These technologies enable real-time processing, interoperability, and modular service design, thereby supporting scalable and adaptive business models.

Cloud computing, for instance, enhances scalability, flexibility, and cost efficiency, while enabling advanced data storage and processing capabilities. APIs facilitate the integration of internal systems with external partners, supporting the development of open banking ecosystems and embedded financial services (Zetzsche et al., 2018). Together, these technologies underpin a shift toward platform-oriented banking models, in which institutions operate not only as service providers but also as orchestrators of multi-sided ecosystems (Parker et al., 2016; Haddad and Hornuf, 2019). In such environments, competitive advantage is increasingly derived from ecosystem positioning and network participation rather than from proprietary products alone.

Importantly, backend infrastructures and data architectures are no longer merely technical enablers but have become strategic assets. By externalising capabilities through APIs and enabling third-party integration, banks can foster innovation, create new revenue streams, and participate in broader financial ecosystems (Gomber et al., 2018). This reflects a reconfiguration of the banking value chain, where value is co-created across organisational boundaries and mediated through digital platforms. From a theoretical perspective, this development aligns with service-dominant logic, which emphasises resource integration and value co-creation among multiple actors (Vargo and Lusch, 2008).

At the same time, increasing reliance on complex digital infrastructures introduces new risks and organisational challenges. The adoption of cloud services and open architectures expands the technological attack surface, raising concerns related to cybersecurity, data privacy, and regulatory compliance (Zetzsche et al., 2018). Moreover, legacy system integration remains a persistent constraint, often resulting in hybrid architectures that limit agility and increase operational complexity (Jöhnk et al., 2021). These challenges highlight that digital maturity is not a linear progression but a path-dependent process, shaped by historical investments, organisational capabilities, and regulatory environments.

Another critical dimension of digital maturity is its relationship with customer experience. As digital services become ubiquitous, differentiation increasingly depends on the ability to deliver seamless, personalised, and context-aware interactions (Lemon and Verhoef, 2016). Achieving this requires real-time data processing, advanced analytics, and integrated customer journeys—capabilities that are fundamentally rooted in backend infrastructure rather than front-end design alone (Wedel and Kannan, 2016). Consequently, digital maturity forms the operational foundation for customer-centric value creation.

However, the relationship between digital maturity and performance is not unambiguously positive. Research on the “digitalisation paradox” suggests that substantial investments in digital technologies do not necessarily translate into immediate productivity gains due to the need for complementary organisational changes and learning effects (Brynjolfsson et al., 2021). In banking, this implies that digital maturity must be accompanied by transformation in organisational structures, governance mechanisms, and corporate culture. Without such alignment, digital initiatives risk remaining fragmented and failing to generate sustained value.

Within the broader framework of this paper, digital maturity represents the foundational layer of transformation. It provides the infrastructural and organisational preconditions upon which more advanced developments—particularly AI-driven intelligence, ecosystem integration, and autonomous financial systems—are built. Without sufficient digital maturity, the deployment of advanced AI capabilities and participation in platform ecosystems remains constrained.

In summary, digital maturity is a multidimensional strategic capability that extends beyond technological adoption to encompass the integration of infrastructure, data, and organisational processes. Its evolution reflects a shift from front-end digitalisation toward backend transformation, enabling platform-based and ecosystem-oriented banking models. While digitally mature institutions are better positioned to compete in increasingly complex and dynamic environments, achieving digital maturity requires continuous adaptation, organisational alignment, and effective management of technological and regulatory risks.

3. Demographic Change and the Transformation of Pension and Wealth Advisory

Demographic change—most notably population ageing and the retirement of the baby boomer generation—constitutes a structural driver of transformation in the banking sector. Across developed economies, increasing life expectancy and declining birth rates are reshaping the demand for financial services, particularly in the domains of retirement planning, decumulation strategies, and intergenerational wealth transfer (Horneff et al., 2013; OECD, 2023). As individuals transition from wealth accumulation to decumulation, financial decision-making becomes significantly more complex, requiring the integration of investment management, tax optimisation, longevity risk assessment, and estate planning.

This growing complexity challenges traditional banking models, which have historically been organised around product-centric offerings. Standardised financial products are insufficient to address the dynamic and interdependent nature of retirement-related financial decisions. As a result, banks are increasingly transitioning toward integrated, advisory-centric service models that provide holistic support across the customer lifecycle. These models combine financial planning, tax advisory, and estate management into unified service propositions, reflecting a shift from transactional interactions to long-term relational engagement (Lemon and Verhoef, 2016).

Empirical evidence suggests that such integrated advisory approaches improve both customer outcomes and institutional performance. Holistic financial planning enables more coherent decision-making under uncertainty, particularly in managing longevity risk and income sustainability during retirement (Horneff et al., 2013). At the same time, these models strengthen customer relationships, increase retention, and enable banks to capture a larger share of wallet, particularly in high-value segments such as affluent and retiree clients.

This transformation aligns with the broader emergence of financial well-being ecosystems, in which financial institutions extend their role beyond traditional intermediation to support customers’ long-term financial security and quality of life (Brüggen et al., 2017). Financial well-being encompasses not only the ability to meet current and future financial obligations but also a sense of control and confidence in financial decision-making. Within this framework, banks act as orchestrators of interconnected services that span multiple domains, including insurance, healthcare planning, and legal advisory, often delivered through partnerships and platform-based models.

From a theoretical perspective, this evolution can be understood through the lens of service-dominant logic, which emphasises value co-creation through resource integration across multiple actors (Vargo and Lusch, 2008). Retirement represents a critical life event around which such resource integration becomes particularly salient. Banks, fintech firms, insurers, and advisory providers increasingly collaborate within ecosystem structures to deliver comprehensive and context-specific solutions tailored to individual needs.

Digital technologies and advanced analytics play a critical enabling role in this transformation. Data-driven insights allow for personalised financial planning, scenario modelling, and dynamic adjustment of investment and withdrawal strategies in response to changing circumstances (Horneff et al., 2006). Robo-advisory platforms and hybrid advisory models—combining algorithmic support with human expertise—have significantly expanded the scalability and accessibility of advisory services (Jung et al., 2018). These developments are particularly important in addressing the growing “advisory gap,” as traditional relationship-based models struggle to meet increasing demand.

Within the framework of this paper, these developments illustrate the interaction between Intelligence and Ecosystems. AI-driven analytics enable the personalisation and scalability of advisory services (Intelligence), while platform-based integration facilitates the coordination of multiple service providers across domains (Ecosystems). Together, these dimensions support the transition toward lifecycle-oriented, customer-centric financial service models.

However, this transformation also introduces important challenges and tensions. The integration of multiple service domains increases operational complexity and raises questions regarding data governance, privacy, and regulatory compliance. Moreover, the reliance on algorithmic decision-making in advisory contexts introduces risks related to model bias, transparency, and suitability of recommendations (Rai, 2020). Ensuring that automated or semi-automated advisory systems align with customers’ best interests remains a critical governance concern, particularly in highly regulated environments.

In addition, competitive dynamics in the advisory space are intensifying. Non-bank actors—including fintech firms, independent wealth platforms, and technology companies—are increasingly entering the market, leveraging superior digital capabilities and user-centric design (Haddad and Hornuf, 2019). These entrants challenge incumbent banks by offering more flexible, accessible, and often lower-cost advisory solutions. In response, banks must differentiate themselves through trust, regulatory expertise, and the ability to deliver integrated, lifecycle-oriented services at scale.

In summary, demographic change acts as a catalyst for the transformation of pension and wealth advisory services. The increasing complexity of retirement-related financial decisions is driving a shift from product-centric models toward integrated, ecosystem-based advisory frameworks. Enabled by digital technologies and advanced analytics, these models support more personalised and scalable service delivery while strengthening long-term customer relationships. At the same time, they introduce new challenges related to governance, competition, and operational complexity, underscoring the need for strategic alignment between technological capabilities, organisational structures, and regulatory requirements.

4. Generative Artificial Intelligence and the Transformation of Banking Operations

Generative artificial intelligence (AI) represents a significant inflection point in the digital transformation of banking operations. While earlier phases of automation were primarily based on deterministic rules and robotic process automation (RPA), generative AI introduces capabilities associated with natural language understanding, contextual reasoning, and dynamic content generation. This shift enables a transition from rule-based execution toward cognitive process automation, in which systems can interpret, generate, and act upon complex and unstructured information (Davenport and Ronanki, 2018; Huang and Rust, 2021).

In operational contexts, generative AI fundamentally alters how information is processed and decisions are supported. Traditional banking processes—particularly in areas such as case management, customer service, and compliance—have historically relied on structured data and predefined workflows. Generative AI systems, by contrast, can ingest and synthesise unstructured data, including text, documents, and conversational inputs, thereby enabling more flexible and context-aware process execution. This significantly expands the scope of automation, allowing systems to handle tasks that previously required human interpretation and judgment (Jöhnk et al., 2021).

Within case management, generative AI enables the automated classification, prioritisation, and resolution of complex inquiries. By analysing historical cases, contextual signals, and real-time inputs, these systems can generate recommendations, draft responses, and support decision-making processes. Empirical evidence suggests that such capabilities improve operational efficiency, reduce processing times, and enhance consistency in decision outcomes (Ryll et al., 2020). Importantly, the value of generative AI in this context lies not only in cost reduction but in its ability to augment cognitive capacity within operational workflows.

In customer service, generative AI facilitates a shift from scripted interactions to adaptive and conversational engagement. Large language model-based systems can handle a wide range of customer queries, from routine requests to more complex, multi-step interactions, while maintaining contextual awareness across conversations (Huang and Rust, 2021). This enhances responsiveness, reduces resolution times, and improves service standardisation. More importantly, it enables a move toward continuous and personalised interaction, where services are dynamically tailored to individual customer contexts.

From a conceptual perspective, these developments reflect a broader transition from automation to augmented intelligence. Rather than replacing human decision-making, generative AI extends human capabilities by supporting analysis, interpretation, and communication in complex environments (Brynjolfsson and McAfee, 2017; Raisch and Krakowski, 2021). In practice, this results in hybrid human–AI workflows, where AI systems perform data-intensive and repetitive tasks, while human agents focus on judgment, exception handling, and relationship management.

Within the framework of this paper, generative AI constitutes the core enabler of the Intelligence dimension. It transforms static, rule-based systems into adaptive, learning infrastructures capable of continuous improvement. Moreover, it provides the cognitive foundation for more advanced developments, including ecosystem integration and autonomous financial systems. In this sense, generative AI does not represent an isolated technological innovation but a structural shift in how banking operations are organised and executed.

However, the deployment of generative AI also introduces significant challenges. Issues of transparency, explainability, and model bias are particularly critical in regulated environments, where decision-making must be auditable and compliant with legal standards (Rai, 2020; Veale and Borgesius, 2021). The probabilistic nature of generative models raises concerns regarding reliability and consistency, particularly in high-stakes contexts such as credit decisions or fraud investigations. Furthermore, the risk of “hallucinations” or factually incorrect outputs necessitates robust validation mechanisms and human oversight.

Data governance and privacy considerations also become more complex. Generative AI systems require access to large volumes of structured and unstructured data, increasing exposure to data protection risks and regulatory scrutiny. Ensuring that these systems operate within ethical and legal boundaries is essential for maintaining customer trust and institutional legitimacy.

In addition, organisational challenges must be addressed. The effective integration of generative AI depends on employee trust, digital literacy, and the redesign of workflows to accommodate human–AI collaboration (Jöhnk et al., 2021). Without such alignment, the benefits of AI adoption may remain limited or unevenly distributed across the organisation.

In summary, generative AI is redefining banking operations by enabling cognitive process automation, enhancing decision-making capabilities, and supporting more adaptive and personalised service delivery. As a core component of the Intelligence dimension, it provides the foundation for broader transformations in banking, including ecosystem integration and the emergence of autonomous financial systems. However, its successful implementation requires careful management of technological, organisational, and governance challenges, particularly in relation to transparency, accountability, and trust.

5. Agentic Commerce and the Emergence of Autonomous Financial Ecosystems

The evolution of artificial intelligence in banking is increasingly extending beyond decision support toward autonomous action. While generative AI enhances cognitive capabilities within organisational processes, a new class of systems—commonly referred to as agentic AI—introduces the capacity to initiate, coordinate, and execute tasks with limited human intervention. This development marks a transition from intelligence augmentation to operational autonomy, fundamentally reshaping how financial transactions are initiated and managed.

Agentic commerce refers to economic interactions in which autonomous or semi-autonomous software agents act on behalf of users to perform financial decisions and transactions. These agents are capable of interpreting user preferences, processing contextual information, and interacting with digital infrastructures to complete tasks such as payments, portfolio rebalancing, subscription management, or procurement decisions. In contrast to traditional automation, which follows predefined rules, agentic systems operate in dynamic environments, adapting their behaviour based on real-time data and learned patterns (Rai, 2020; Gomber et al., 2018).

In the context of banking, agentic systems have the potential to redefine the interface between customers and financial institutions. Rather than directly engaging with banking platforms, users may increasingly rely on intelligent agents that mediate interactions, select financial products, and execute transactions autonomously. This shifts the locus of control from user-driven interfaces to algorithmically mediated decision environments, where financial choices are increasingly embedded in automated processes.

This transformation has significant implications for the structure of financial markets. As agentic systems interact with one another across platforms, they give rise to autonomous financial ecosystems characterised by machine-to-machine (M2M) transactions and continuous, real-time coordination. In such ecosystems, financial services become embedded within broader digital infrastructures, including e-commerce platforms, smart devices, and Internet of Things (IoT) networks. Payments, lending decisions, and investment actions may occur seamlessly and invisibly as part of integrated service flows, rather than as discrete, user-initiated events.

From a theoretical perspective, agentic commerce represents a further extension of platform and ecosystem theories. While existing research has emphasised the role of digital platforms in facilitating multi-sided interactions (Parker et al., 2016), agentic systems introduce a new layer of automation in which interactions are no longer solely human-driven. Value creation increasingly occurs through the coordination of autonomous agents operating across organisational boundaries. This reinforces the shift from product-centric to ecosystem-based models, while simultaneously introducing new forms of intermediation and control.

Within the IEA framework developed in this paper, agentic commerce constitutes the core of the Autonomy dimension. It builds on the Intelligence layer—enabled by AI-driven data processing and decision-making—and extends it toward the execution of actions without direct human input. At the same time, it is inherently linked to the Ecosystems dimension, as autonomous agents rely on interoperable infrastructures, APIs, and platform-based architectures to operate effectively. Autonomy, therefore, emerges not as an isolated capability but as the outcome of the interaction between intelligence and ecosystem integration.

Despite its transformative potential, agentic commerce introduces substantial challenges. One critical issue concerns control and accountability. As decision-making authority shifts from humans to algorithms, questions arise regarding responsibility for outcomes, particularly in cases of financial loss, erroneous transactions, or unintended behaviour. The opacity of complex AI systems further complicates these issues, making it difficult to trace and explain decision processes (Rai, 2020).

Trust also becomes a central concern. Users must be willing to delegate financial decision-making authority to autonomous agents, which requires confidence in system reliability, security, and alignment with user preferences. However, the potential for algorithmic errors, adversarial manipulation, or misaligned incentives may undermine this trust. Ensuring robust governance mechanisms, including transparency, auditability, and fail-safe controls, is therefore essential.

In addition, the rise of agentic systems has significant competitive implications. As intermediaries between users and financial institutions, autonomous agents may assume a gatekeeping role, influencing which services are selected and how transactions are executed. This creates the potential for new forms of market power, particularly for technology firms that control the underlying AI infrastructures or user interfaces. Consequently, competition in banking may increasingly shift from product differentiation to algorithmic visibility and platform positioning, a dynamic further explored in later sections of this paper.

Regulatory frameworks are also challenged by these developments. Existing financial regulations are largely designed for human-driven decision-making and discrete transactions. The emergence of continuous, automated, and cross-platform interactions complicates oversight, requiring new approaches to supervision, risk management, and consumer protection (Zetzsche et al., 2018). In particular, regulators must address issues related to algorithmic accountability, systemic risk arising from interconnected agents, and the protection of user autonomy.

In summary, agentic commerce represents a critical step in the evolution of banking toward autonomous financial systems. By enabling software agents to initiate and execute transactions on behalf of users, it transforms the nature of financial interaction and redefines the boundaries of financial institutions. As the core of the Autonomy dimension within the IEA framework, agentic systems build on AI-driven intelligence and ecosystem integration to create a new paradigm of financial service delivery. However, their adoption raises fundamental challenges related to control, trust, competition, and regulation, underscoring the need for new governance models in increasingly autonomous financial environments.

6. AI-Enabled Workplace Transformation

The integration of artificial intelligence into banking is not limited to customer-facing services and operational processes; it is fundamentally transforming the nature of work within financial institutions. In particular, AI is reshaping knowledge work, altering how information is processed, decisions are made, and organisational tasks are coordinated. This transformation extends beyond automation to redefine the relationship between human expertise and technological capability.

Historically, banking has relied heavily on highly skilled professionals to perform complex analytical and decision-intensive tasks, including credit assessment, risk management, compliance analysis, and financial advisory. These activities are characterised by the interpretation of large volumes of information, the application of domain expertise, and the exercise of judgment under uncertainty. The introduction of AI—particularly generative and predictive systems—significantly alters this paradigm by enabling machines to perform substantial parts of these cognitive tasks (Brynjolfsson et al., 2021; Raisch and Krakowski, 2021).

AI systems can now synthesise large and diverse datasets, generate insights, and support decision-making processes in real time. For example, in credit analysis, AI can integrate structured financial data with unstructured information such as text-based reports or market signals to produce more comprehensive risk assessments. In compliance functions, AI can assist in monitoring transactions, identifying anomalies, and interpreting regulatory requirements. These capabilities enhance both the speed and scope of analysis, allowing organisations to process information at a scale that exceeds human capacity.

From a conceptual perspective, this transformation reflects a shift from task automation to cognitive augmentation. Rather than replacing human workers, AI increasingly acts as a collaborative agent that complements human expertise. Hybrid human–AI systems enable a division of labour in which machines perform data-intensive, repetitive, and pattern-recognition tasks, while humans focus on interpretation, ethical judgment, and relationship management (Raisch and Krakowski, 2021). This reconfiguration of work has significant implications for productivity, organisational design, and skill requirements.

Empirical research suggests that AI adoption can lead to substantial productivity gains, particularly in knowledge-intensive sectors. By reducing the time required for information processing and analysis, AI enables employees to focus on higher-value activities, thereby increasing overall efficiency (Brynjolfsson et al., 2021). However, these gains are not automatic. Realising the benefits of AI requires complementary investments in organisational processes, employee training, and workflow redesign. Without such adjustments, AI systems may be underutilised or fail to integrate effectively into existing work structures.

The transformation of the workplace also entails significant changes in skill requirements. Demand is increasing for employees who can work effectively with AI systems, interpret algorithmic outputs, and oversee automated processes. This includes not only technical skills but also critical thinking, domain expertise, and the ability to evaluate and challenge AI-generated insights. Consequently, continuous learning and reskilling become essential components of organisational strategy.

Within the IEA framework, these developments reinforce the Intelligence dimension by embedding AI capabilities directly into organisational processes and decision-making structures. At the same time, they form a critical precondition for the emergence of the Autonomy dimension. As organisations become more reliant on AI-supported decision-making, the boundary between decision support and autonomous action becomes increasingly fluid. In this sense, the transformation of the workplace represents a transitional stage between intelligence augmentation and operational autonomy.

Despite its potential, the integration of AI into the workplace introduces several challenges. One key issue concerns trust and reliance. Employees must develop confidence in AI systems while maintaining the ability to critically assess their outputs. Overreliance on AI may lead to automation bias, where users accept algorithmic recommendations without sufficient scrutiny, while underreliance may limit the benefits of AI adoption (Rai, 2020).

Another important concern relates to transparency and explainability. In many cases, AI systems—particularly those based on complex machine learning models—operate as “black boxes,” making it difficult for users to understand how decisions are generated. This lack of transparency can hinder adoption, complicate accountability, and raise regulatory concerns, particularly in highly regulated environments such as banking (Veale and Borgesius, 2021).

Organisational implications must also be considered. The integration of AI may lead to shifts in job roles, changes in hierarchical structures, and the emergence of new forms of coordination between human and machine actors. While some tasks may be displaced, others are likely to be transformed or newly created, resulting in a reconfiguration rather than a simple reduction of work. Managing this transition requires careful change management, clear governance structures, and alignment between technological and organisational strategies.

Finally, ethical considerations play an increasingly important role. The use of AI in decision-making raises questions regarding fairness, bias, and accountability, particularly when decisions have significant financial or social consequences. Ensuring that AI systems operate in a transparent, fair, and responsible manner is essential for maintaining trust among employees, customers, and regulators.

In summary, artificial intelligence is fundamentally transforming the banking workplace by reshaping knowledge work, enhancing cognitive capabilities, and redefining the relationship between human and machine actors. As a core component of the Intelligence dimension, AI enables more efficient and scalable decision-making processes while creating the organisational foundations for increased autonomy. However, its successful integration depends on complementary investments in skills, governance, and organisational design, as well as careful management of risks related to trust, transparency, and ethics.

7. Customer-Centric Operating Models

Customer-centricity has become a defining principle of contemporary banking transformation. However, in increasingly digital and AI-driven environments, customer-centricity extends beyond improved service quality or user experience design. It reflects a fundamental shift in the operating model of financial institutions, in which products, processes, and organisational structures are reconfigured around continuously evolving customer needs and contexts.

Traditionally, banking operating models have been organised around product silos, such as lending, payments, and wealth management. This structure prioritised internal efficiency and product optimisation but often resulted in fragmented customer experiences and limited integration across service domains. In contrast, customer-centric operating models are organised around customer journeys and lifecycle events, enabling institutions to deliver more coherent and context-aware services (Lemon and Verhoef, 2016).

Digital technologies—and particularly artificial intelligence—play a central role in enabling this transformation. AI allows financial institutions to analyse large volumes of customer data, identify behavioural patterns, and generate real-time insights into customer preferences and needs. These capabilities support the delivery of personalised and adaptive services, where product offerings, communication, and interactions are dynamically tailored to individual users (Wedel and Kannan, 2016). As a result, customer engagement shifts from reactive service provision toward proactive and predictive value creation.

From a conceptual perspective, this transformation reflects the integration of the Intelligence and Ecosystems dimensions of the IEA framework. The Intelligence dimension provides the analytical capabilities required to understand and anticipate customer behaviour, while the Ecosystems dimension enables the integration of services across organisational boundaries. Together, these dimensions support the development of end-to-end customer solutions that extend beyond the traditional scope of banking.

Ecosystem integration is particularly important in this context. Financial needs are often embedded in broader life events, such as purchasing a home, planning for retirement, or managing healthcare expenses. Addressing these needs requires the coordination of multiple services, including insurance, real estate, legal advisory, and investment management. Platform-based models and API-driven architectures allow banks to integrate these services into unified customer journeys, thereby enhancing convenience and value creation (Parker et al., 2016; Zetzsche et al., 2018).

As customer-centric operating models evolve, the role of the bank shifts from that of a product provider to an orchestrator of customer ecosystems. Competitive advantage increasingly depends on the ability to coordinate and integrate services, manage customer relationships across channels, and leverage data to deliver personalised experiences at scale. This shift also alters the nature of competition, as banks must compete not only with traditional financial institutions but also with fintech firms and technology platforms that offer integrated, user-centric solutions (Haddad and Hornuf, 2019).

Artificial intelligence further amplifies these dynamics by enabling continuous and context-aware interaction. Customer engagement is no longer limited to discrete touchpoints but becomes an ongoing process in which services are dynamically adjusted based on real-time data. For example, AI systems can anticipate financial needs, recommend actions, and automate routine decisions, thereby reducing friction and enhancing user experience. In this sense, customer-centricity increasingly overlaps with elements of autonomy, as certain decisions are delegated to AI systems operating on behalf of the customer.

However, the implementation of customer-centric operating models also introduces significant challenges. One key issue concerns data privacy and trust. The delivery of personalised services relies on extensive data collection and analysis, raising concerns about how customer data is used, stored, and shared. Maintaining transparency and ensuring compliance with data protection regulations are essential for sustaining customer trust.

Another challenge relates to over-personalisation and algorithmic bias. While personalisation can enhance relevance and engagement, it may also lead to unintended consequences, such as reinforcing existing behavioural patterns or limiting customer choice. Ensuring that AI-driven recommendations remain fair, transparent, and aligned with customer interests is therefore a critical governance issue (Rai, 2020).

In addition, ecosystem-based models create dependencies on external partners and platforms. While such partnerships enable innovation and service integration, they also introduce risks related to coordination, control, and value capture. Banks must carefully manage these relationships to avoid losing strategic positioning within the ecosystem.

From an organisational perspective, the transition to customer-centric operating models requires significant structural changes. This includes the breakdown of product silos, the adoption of cross-functional teams organised around customer journeys, and the implementation of agile processes that support rapid adaptation. It also requires alignment between technological capabilities, data governance, and strategic objectives.

In summary, customer-centric operating models represent a fundamental transformation in how banks organise and deliver financial services. Enabled by AI and digital ecosystems, these models shift the focus from products to customer journeys, from reactive service provision to proactive value creation, and from isolated offerings to integrated solutions. As a key interface between technological capabilities and market outcomes, customer-centricity plays a central role in linking the Intelligence and Ecosystems dimensions of the IEA framework, while also laying the groundwork for increasing levels of autonomy in financial decision-making. However, its successful implementation depends on the effective management of data, trust, and ecosystem relationships in an increasingly complex and competitive environment.

8. Emerging Risks: Deepfakes and Digital Identity

The increasing reliance on digital technologies and artificial intelligence in banking has significantly expanded the importance of digital identity as a foundational component of financial systems. At the same time, it has introduced new categories of risk, particularly those associated with synthetic media, deepfakes, and AI-enabled fraud. These developments challenge existing assumptions about trust, authentication, and the integrity of digital interactions, thereby raising fundamental questions for financial institutions and regulators.

Digital identity can be understood as the set of attributes, credentials, and behavioural signals that enable the identification and verification of individuals in digital environments. In traditional banking, identity verification relied on physical documentation and in-person interactions. However, the shift toward digital channels has necessitated the development of remote and automated identity verification mechanisms, including biometric authentication, document recognition, and behavioural analytics (Zetzsche et al., 2018). As a result, digital identity has evolved from a procedural requirement into a core infrastructural element of banking operations.

Within AI-driven systems, digital identity plays an even more critical role. As financial services become increasingly personalised, automated, and interconnected, accurate and secure identification is essential for enabling transactions, managing risk, and ensuring regulatory compliance. In the context of the IEA framework, digital identity acts as an enabling layer that supports both the Intelligence dimension—by providing reliable data inputs—and the Autonomy dimension, where systems may initiate and execute actions on behalf of users. Without robust identity mechanisms, the expansion of autonomous financial processes becomes inherently constrained.

At the same time, advances in artificial intelligence have significantly increased the sophistication of identity-related threats. Deepfakes and synthetic media—generated using machine learning techniques—allow malicious actors to create highly realistic but fabricated audio, video, and text content. These technologies can be used to impersonate individuals, bypass authentication systems, and manipulate financial transactions (Chesney and Citron, 2019). In banking, such threats are particularly concerning, as they directly target trust-based processes such as identity verification, authorisation, and customer communication.

The implications of deepfake technologies extend beyond isolated fraud incidents. They undermine the reliability of digital signals that underpin AI-driven decision-making systems. For example, if biometric authentication mechanisms can be spoofed or manipulated, the integrity of identity verification processes is compromised. Similarly, AI-generated communications may be used to deceive both customers and employees, increasing the risk of social engineering attacks and operational vulnerabilities.

These developments highlight a broader structural challenge: as AI enhances both defensive and offensive capabilities, the security landscape becomes increasingly dynamic and adversarial. Financial institutions must therefore adopt more sophisticated approaches to risk management, including the use of AI-based detection systems, multi-factor authentication, and continuous monitoring of behavioural patterns. In this context, security evolves from a static control mechanism to an adaptive and intelligence-driven process.

However, technological solutions alone are insufficient. The growing complexity of identity systems raises important governance and regulatory questions. Ensuring the integrity, privacy, and portability of digital identities requires coordinated efforts across institutions, industries, and jurisdictions. Emerging frameworks for digital identity—such as decentralised identity models and self-sovereign identity—seek to provide individuals with greater control over their data while maintaining security and interoperability (Zetzsche et al., 2018). Nevertheless, their implementation remains uneven and raises additional challenges related to standardisation, scalability, and user adoption.

Trust emerges as a central theme in this context. The effectiveness of digital banking systems depends on the confidence of users that their identities are protected and that transactions are secure. However, the proliferation of AI-generated content and increasingly sophisticated fraud techniques may erode this trust. Maintaining trust therefore requires not only robust technological safeguards but also transparency, clear communication, and effective regulatory oversight.

From a competitive perspective, the ability to manage identity-related risks may become a source of differentiation. Institutions that can provide secure, reliable, and user-friendly identity solutions are better positioned to participate in digital ecosystems and support autonomous financial services. Conversely, failures in identity management can have significant reputational and financial consequences, undermining both customer confidence and regulatory compliance.

In summary, digital identity represents a foundational pillar of AI-driven banking, enabling secure and reliable interactions in increasingly digital and automated environments. At the same time, advances in AI—particularly in the form of deepfakes and synthetic media—introduce new and evolving risks that challenge traditional approaches to authentication and trust. Addressing these challenges requires a combination of advanced technological solutions, robust governance frameworks, and coordinated regulatory efforts. As banking systems continue to evolve toward greater intelligence and autonomy, the effective management of digital identity will be essential for ensuring security, trust, and long-term sustainability.

9. AI Visibility and the Future of Banking Competition

A new competitive dimension is emerging in banking through AI-mediated information environments that increasingly shape how customers discover, evaluate, and select financial services. As generative AI systems, virtual assistants, and recommendation engines become primary interfaces for financial information, AI visibility is becoming a strategic determinant of market positioning. AI visibility refers to the extent to which a bank’s products, services, and brand are accurately represented within algorithmically generated outputs, thereby influencing customer perception prior to any direct institutional interaction (Hundertmark et al., 2024).

This development extends beyond traditional search engine optimization. While SEO focused on improving visibility in ranked search results, AI visibility reflects a shift toward algorithmic mediation of meaning, where systems synthesize, interpret, and recommend financial information based on inferred relevance, credibility, and contextual alignment (Kaplan and Haenlein, 2020). As a result, banks no longer compete solely for customer attention but for algorithmic interpretability and representational accuracy within AI systems.

From a strategic perspective, AI visibility is directly shaped by data architecture and information governance. Banks with structured, high-quality, and machine-readable data are more likely to be correctly interpreted and favorably represented by AI systems. Conversely, fragmented or inconsistent data increases the risk of distortion, omission, or mischaracterization in AI-generated outputs. In regulated environments, such distortions are not merely reputational risks but may also produce compliance and legal exposure (Rai, 2020).

This introduces a shift in competitive logic. Traditional performance indicators—such as market share, digital traffic, or brand awareness—become insufficient predictors of customer acquisition in AI-mediated environments. Instead, competitive positioning increasingly depends on how institutions are encoded within foundation models and recommendation systems, including the accuracy, prominence, and framing of their offerings (Sorescu et al., 2020). In this context, banks effectively compete within an informational layer that is partially external to their direct control.

This shift also introduces new governance imperatives. Ensuring AI visibility requires continuous investment in data standardization, semantic consistency, and transparent product representation. At the same time, banks must engage with emerging AI ecosystems to monitor how their services are interpreted and presented across different models and platforms. Misalignment between institutional intent and AI output becomes a strategic vulnerability, particularly as customers increasingly rely on AI systems as primary decision intermediaries.

At the same time, AI visibility introduces structural uncertainty. Generative systems may reproduce biases, hallucinate product features, or simplify complex financial offerings in ways that distort institutional reality. This creates a tension between algorithmic efficiency and informational fidelity, particularly in domains where precision and regulatory compliance are critical.

In conclusion, AI visibility represents a structural transformation in banking competition. Competitive advantage is increasingly determined not only by product quality or digital presence, but by how effectively institutions are represented within AI-mediated decision environments. Managing this requires a combination of robust data governance, algorithmic literacy, and active participation in shaping how financial information is encoded within emerging AI ecosystems.

10. Next Steps in Banking Digitalization

The developments discussed in this paper converge into three interrelated structural shifts that define the next phase of banking digitalization.

First, banking is moving from digitalization to intelligence. Early transformation focused on digitizing processes and interfaces; the next phase is defined by AI systems that actively support, augment, and increasingly automate decision-making. Intelligence becomes embedded within operational and customer-facing systems rather than layered on top of them.

Second, the industry is shifting from products to ecosystems. Financial services are increasingly embedded within broader customer journeys and life events, rather than delivered as standalone offerings. Value creation depends less on isolated products and more on integration across platforms, partners, and service layers. Banking becomes a component of interconnected digital ecosystems rather than a discrete service domain.

Third, banking is transitioning from human-controlled to partially autonomous systems. Agentic AI introduces systems capable of executing tasks, initiating actions, and managing financial processes with limited human intervention. This does not imply full autonomy but rather the gradual redistribution of decision authority between humans and machine agents.

These shifts are mutually reinforcing. Intelligence enables ecosystem orchestration, ecosystems generate the data required for further intelligence, and autonomy operationalizes both at scale. Together, they point toward a future in which banking becomes increasingly embedded, anticipatory, and operationally invisible to end users.

At the same time, this trajectory introduces important constraints and tensions. Increased autonomy raises questions of accountability and oversight. Ecosystem dependence creates exposure to platform concentration and third-party governance. Intelligence-driven systems introduce risks related to explainability, bias, and model uncertainty. These tensions suggest that digitalization is not a linear progression toward efficiency, but a structural reconfiguration of control, transparency, and responsibility in financial systems.

In sum, the next phase of banking digitalization is defined not simply by technological adoption, but by a deeper reorganization of how financial services are produced, delivered, and governed across increasingly intelligent and autonomous digital environments.

11. Theory Development: The IEA Framework of AI-Driven Banking Transformation

This paper develops the IEA Framework (Intelligence–Ecosystems–Autonomy) as a conceptual model to explain the structural transformation of banking under conditions of artificial intelligence, platformization, and advanced automation. The framework integrates insights from digital transformation theory, platform economics, and artificial intelligence research to conceptualize banking as an evolving system of interdependent technological and organizational architectures, rather than a sequence of isolated innovations.

The core argument is that contemporary banking transformation is not driven by discrete technologies in isolation, but by the co-evolution of cognitive capability, network structure, and decision autonomy. The IEA Framework therefore positions intelligence, ecosystems, and autonomy as mutually constitutive dimensions of a unified transformation process that reshapes value creation, organizational design, and competitive logic.

11.1 Intelligence: From Digitization to Cognitive Financial Systems

The Intelligence dimension captures the shift from rule-based digital systems toward AI-enabled cognitive infrastructures capable of learning, inference, and decision augmentation. While early digital transformation focused on digitization and workflow automation, contemporary AI systems introduce a qualitative shift toward context-sensitive, data-driven organizational cognition (Brynjolfsson and McAfee, 2017; Huang and Rust, 2021).

In banking, intelligence is operationalized through real-time fraud detection, predictive credit scoring, personalized advisory systems, and conversational interfaces. These applications increasingly move beyond descriptive and diagnostic analytics toward prescriptive and generative decision support, enabling institutions to anticipate needs and adapt responses dynamically (Raisch and Krakowski, 2021).

Crucially, intelligence in this framework is not merely an efficiency enhancement but a reconfiguration of organizational cognition, where decision-making is distributed across human and machine systems. However, intelligence is structurally constrained. Issues of interpretability, bias, and epistemic opacity limit the reliability of algorithmic outputs in high-stakes financial contexts. As intelligence increases, so does the need for governance, explainability, and control, making intelligence simultaneously a capability and a constraint within financial systems (Rai, 2020).

11.2 Ecosystems: From Firm-Centric Banking to Platform-Oriented Finance

The Ecosystems dimension describes the transition from vertically integrated banking institutions toward distributed, platform-based financial architectures. In this configuration, value creation is no longer contained within firm boundaries but is distributed across networks of banks, fintech firms, infrastructure providers, and third-party developers (Parker et al., 2016; Gomber et al., 2018).

Open banking regimes, API-based architectures, and embedded finance accelerate this shift by enabling standardized interoperability across institutional boundaries. Consequently, banks increasingly operate as orchestrators of financial ecosystems rather than closed service providers.

This transformation reflects a move toward service-dominant logic, where value emerges through resource integration and co-creation across multiple actors (Vargo and Lusch, 2008). Competitive advantage is therefore increasingly determined by ecosystem position, data access, and platform control, rather than proprietary product ownership alone.

However, ecosystem participation introduces structural dependencies. Interoperability requirements, governance fragmentation, and reliance on external platforms create new forms of strategic exposure, particularly in relation to data ownership and infrastructural control.

11.3 Autonomy: Toward Self-Executing Financial Systems

The Autonomy dimension captures the emergence of financial systems capable of executing actions with limited or no human intervention. While traditional automation focuses on task execution, autonomy embeds decision logic directly into software agents and infrastructure layers, enabling systems to initiate and complete financial actions based on objectives, preferences, and learned behaviour.

In banking, autonomy manifests in algorithmic trading, robo-advisory systems, automated payment execution, and emerging agent-based financial infrastructures capable of end-to-end transaction management (Jung et al., 2018; Rai, 2020).

Recent advances in generative and agentic AI extend this trajectory by enabling systems that can plan, coordinate, and act across digital environments. This signals a shift from human-in-the-loop architectures to human-on-the-loop governance models, where oversight replaces direct operational control (Raisch and Krakowski, 2021). Autonomy therefore represents a structural reallocation of decision rights from institutions and individuals to algorithmic systems embedded within financial infrastructures.

However, autonomy introduces fundamental governance tensions, including accountability gaps, systemic risk propagation, and regulatory misalignment. As decision authority becomes distributed across autonomous agents, traditional notions of responsibility and control become increasingly difficult to operationalize.

11.4 Integration of the IEA Dimensions

The three dimensions of the IEA Framework are deeply interdependent and evolve through recursive feedback loops.

Intelligence enables ecosystems by improving coordination, prediction, and real-time decision capabilities across distributed actors. Ecosystems, in turn, enhance intelligence by generating richer, more diverse datasets that improve model performance and adaptability. Autonomy builds upon both dimensions by embedding intelligence-driven decisions within ecosystem infrastructures, enabling scalable execution without continuous human intervention.

This interdependence generates a self-reinforcing transformation dynamic, in which advances in one dimension accelerate developments in the others. Banking transformation is therefore best understood as an architectural evolution of coupled systems, rather than a linear process of technological adoption.

11.5 Theoretical Propositions

The IEA Framework advances the following core proposition:

P1: Banking transformation is driven by the co-evolution of intelligence, ecosystem integration, and autonomy rather than by any single technological driver.

From this, three interrelated propositions follow:

• P1a: Increases in AI-driven intelligence enhance coordination efficiency within financial ecosystems.

• P1b: Greater ecosystem integration increases the learning capacity and performance of AI systems.

• P1c: The emergence of autonomy is contingent upon the joint maturity of intelligence and ecosystem architectures.

Together, these propositions conceptualize banking transformation as a systemic process of architectural reconfiguration, rather than incremental technological improvement.

11.6 Theoretical Contribution

The IEA Framework contributes to the literature in three ways.

First, it integrates previously fragmented research streams on digital transformation, artificial intelligence, and platform economics into a unified structural model of banking evolution.

Second, it extends existing theory by introducing autonomy as a distinct analytical dimension, moving beyond traditional automation-centric interpretations of digital transformation.

Third, it reconceptualizes banking transformation as a multi-layered architectural process, in which cognitive capability, network structure, and decision authority co-evolve and mutually reinforce one another.

12. Conclusion

This paper set out to develop a structured conceptual understanding of how artificial intelligence is transforming the banking industry beyond incremental digitalization. Through the Intelligence–Ecosystems–Autonomy (IEA) framework, it has argued that the current transformation is not primarily technological in a narrow sense, but architectural in nature, involving a deep restructuring of how financial institutions generate intelligence, organize relationships, and distribute decision authority.

The Intelligence dimension demonstrates that AI is no longer merely a tool for automation but a core component of organizational cognition in banking. Decision-making is increasingly augmented by predictive, generative, and adaptive systems that reshape how banks interpret data, assess risk, and interact with customers. However, this shift also introduces new dependencies on model transparency, data quality, and algorithmic governance.

The Ecosystems dimension highlights that banking is no longer confined to institutional boundaries. Instead, value creation is increasingly distributed across interconnected platforms, APIs, and partnerships that redefine the role of the bank as an orchestrator rather than a sole provider of financial services. While this enables scalability and innovation, it also introduces structural dependencies and governance complexity across multi-actor environments.

The Autonomy dimension captures the most transformative shift: the gradual emergence of financial systems capable of executing decisions independently of direct human control. This evolution challenges traditional governance structures by redistributing decision rights across algorithmic agents, raising fundamental questions about accountability, oversight, and systemic resilience.

Taken together, the IEA framework suggests that banking is entering a phase of co-evolutionary transformation, where intelligence, ecosystems, and autonomy reinforce each other in continuous feedback loops. This process does not merely enhance efficiency but fundamentally alters the architecture of financial systems, including how value is created, how decisions are made, and how institutions compete.

At the same time, this transformation introduces unresolved tensions. Increasing autonomy complicates regulatory oversight. Ecosystem interdependence increases exposure to platform risk. Expanding reliance on AI systems raises concerns regarding bias, interpretability, and operational robustness. These tensions indicate that the future of banking will not be defined solely by technological capability, but by the ability of institutions to govern increasingly complex socio-technical systems.

In conclusion, this paper positions AI not as an incremental innovation within banking, but as a structural force reshaping the foundations of financial intermediation. The IEA framework provides a lens through which this transformation can be understood, analyzed, and further developed in future empirical and theoretical research.

References

Alt, R., Beck, R. and Smits, M.T. (2018) ‘FinTech and the transformation of the financial industry’, Electronic Markets, 28(3), pp. 235–243.

Baber, H. (2020) ‘FinTech, Crowdfunding and Customer Retention in Islamic Banks’, Vision, 24(3), pp. 260–268.

Boot, A.W.A., Hoffmann, P., Laeven, L. and Ratnovski, L. (2021) ‘Fintech: What’s old, what’s new?’, Journal of Financial Stability, 53, 100836.

Brüggen, E.C., Hogreve, J., Holmlund, M., Kabadayi, S. and Löfgren, M. (2017) ‘Financial well-being: A conceptualization and research agenda’, Journal of Business Research, 79, pp. 228–237.

Brynjolfsson, E. and McAfee, A. (2017) Machine, Platform, Crowd. New York: W.W. Norton & Company.

Brynjolfsson, E., Rock, D. and Syverson, C. (2021) ‘The productivity J-curve: How intangibles complement general purpose technologies’, American Economic Journal: Macroeconomics, 13(1), pp. 333–372.

Brynjolfsson, E., Rock, D. and Syverson, C. (2021) ‘The productivity J-curve: How intangibles complement general purpose technologies’, American Economic Journal: Macroeconomics, 13(1), pp. 333–372.

Davenport, T.H., Guha, A., Grewal, D. and Bressgott, T. (2020) ‘How artificial intelligence will change the future of marketing’, Journal of the Academy of Marketing Science, 48(1), pp. 24–42.

Davenport, T.H. and Kirby, J. (2016) Only Humans Need Apply: Winners and Losers in the Age of Smart Machines. New York: Harper Business.

Deloitte (2024) Digital Banking Maturity Study 2024: Swiss banks continue to fall behind in the digitalisation race. Zurich: Deloitte

Gomber, P., Koch, J.A. and Siering, M. (2018) ‘Digital Finance and FinTech: current research and future research directions’, Journal of Business Economics, 87(5), pp. 537–580.

Gomber, P., Kauffman, R.J., Parker, C. and Weber, B.W. (2018) ‘On the fintech revolution: Interpreting the forces of innovation, disruption, and transformation in financial services’, Journal of Management Information Systems, 35(1), pp. 220–265.

Haddad, C. and Hornuf, L. (2019) ‘The emergence of the global fintech market: Economic and technological determinants’, Small Business Economics, 53(1), pp. 81–105.

Horneff, W.J., Maurer, R.H., Mitchell, O.S. and Rogalla, R. (2013) ‘Optimal life cycle portfolio choice with variable annuities offering liquidity and investment flexibility’, Review of Finance, 24(4), pp. 1013–1054.

Huang, M.-H. and Rust, R.T. (2021) ‘A strategic framework for artificial intelligence in marketing’, Journal of the Academy of Marketing Science, 49(1), pp. 30–50.

Jain, A.K., Ross, A. and Nandakumar, K. (2021) Introduction to Biometrics. 2nd edn. Cham: Springer.

Jarrahi, M.H. (2018) ‘Artificial intelligence and the future of work: Human-AI symbiosis in organizational decision making’, Business Horizons, 61(4), pp. 577–586.

Jöhnk, J., Weißert, M. and Wyrtki, K. (2021) ‘Ready or not, AI comes—An interview study of organizational AI readiness factors’, Business & Information Systems Engineering, 63(1), pp. 5–20

Jung, D., Glaser, F. and Köpplin, W. (2018) ‘Robo-advisory: Digitalization and automation of financial advisory services’, Business & Information Systems Engineering, 60(1), pp. 81–86.

Kaplan, A. and Haenlein, M. (2020) ‘Rulers of the world, unite! The challenges and opportunities of artificial intelligence’, Business Horizons, 63(1), pp. 37–50

Kietzmann, J., Lee, L.W., McCarthy, I.P. and Kietzmann, T.C. (2020) ‘Deepfakes: Trick or treat?’, Business Horizons, 63(2), pp. 135–146.

Lemon, K.N. and Verhoef, P.C. (2016) ‘Understanding customer experience throughout the customer journey’, Journal of Marketing, 80(6), pp. 69–96.

McKinsey & Company (2023) The State of AI in 2023: Generative AI’s Breakout Year. Available at: https://www.mckinsey.com

Parker, G.G., Van Alstyne, M.W. and Choudary, S.P. (2016) Platform Revolution: How Networked Markets Are Transforming the Economy. New York: W.W. Norton & Company.

Nguyen, T.T., Nguyen, C.M., Nguyen, D.T., Nguyen, D.T., Nahavandi, S. and Yang, H. (2019) ‘Deep learning for deepfakes creation and detection: A survey’, arXiv preprint arXiv:1909.11573.

Nicoletti, B. (2017) The Future of FinTech: Integrating Finance and Technology in Financial Services. Cham: Springer.

OECD (2023) Pensions at a Glance 2023: OECD and G20 Indicators. Paris: OECD Publishing.

Rai, A. (2020) ‘Explainable AI: From black box to glass box’, Journal of the Academy of Marketing Science, 48(1), pp. 137–141.

Raisch, S. and Krakowski, S. (2021) ‘Artificial intelligence and management: The automation–augmentation paradox’, Academy of Management Review, 46(1), pp. 192–210.

Rigby, D.K., Sutherland, J. and Takeuchi, H. (2016) ‘Embracing agile’, Harvard Business Review, 94(5), pp. 40–50.

Shah, D., Rust, R.T., Parasuraman, A., Staelin, R. and Day, G.S. (2006) ‘The path to customer centricity’, Journal of Service Research, 9(2), pp. 113–124.

Sorescu, A., Sorescu, S.M. and Saunderson, A. (2020) ‘Artificial intelligence and the future of marketing: Opportunities and challenges’, Journal of the Academy of Marketing Science, 48(1), pp. 14–26.

Vargo, S.L. and Lusch, R.F. (2008) ‘Service-dominant logic: continuing the evolution’, Journal of the Academy of Marketing Science, 36(1), pp. 1–10.

Vial, G. (2019) ‘Understanding digital transformation: A review and a research agenda’, The Journal of Strategic Information Systems, 28(2), pp. 118–144.

Wedel, M. and Kannan, P.K. (2016) ‘Marketing analytics for data-rich environments’, Journal of Marketing, 80(6), pp. 97–121.

Zetzsche, D.A., Buckley, R.P., Arner, D.W. and Barberis, J.N. (2018) ‘From FinTech to TechFin: The regulatory challenges of data-driven finance’, New York University Journal of Law & Business, 14(2), pp. 393–446.