Artificial Intelligence and Finance: A bibliometric review on the Trends, Influences, and Research Directions

2.1 Data and methodology

We employed a mixed review technique in this study, combining the TCCM framework with bibliometric review to identify research trends and propose future possibilities within the research context. The TCCM framework examines existing theoretical and research frameworks and suggests changes or the creation of new models (Paul & Rosado-Serrano, 2019). This approach is consistent with the design and content of the accepted review procedures (Bhukya & Paul, 2023). Furthermore, we incorporated inclusion and exclusion criteria (Bhukya & Paul, 2023) when choosing the materials for analysis. Articles from peer-reviewed journals in the Web of Science (WoS) database with an impact factor of at least one were included.

The selection follows a three-step process: i) identification of relevant documents, ii) screening of documents, and iii) final selection. This establishes the relevance and rigour of the materials required for the study. The theories, constructions, contexts, and methodologies in the literature were evaluated using the TCCM framework. Bibliometric analysis reveals information on cocitation networks, publishing trends, and citation patterns. We applied an integrated approach, that help us to uncover new topics and knowledge gaps for future research. This provides a broad picture of the evolution of AI applications in finance. This would establish a foundation for future theoretical and methodological advancements and provide useful recommendations for academics and specialists in the area.

2.2 Identification

The WoS bibliometric database (http://www.webofknowledge.com) was used to identify research articles relevant to AI and finance. The search terms included “artificial intelligence” “AI” and “finance” finance, which were applied to conduct title, abstract, and keyword searches. The initial query retrieved a total of n = 2476 documents connected to AI and finance research. Recognizing the possibility that not all documents would be directly relevant to the research scope, further screening was performed using predefined filters to ensure the selection of pertinent documents.

2.3 Screening

A multilayered filtering process was employed to refine the document set and maintain its relevance to the study objectives. First, the scope was narrowed to publications categorized under the fields of ‘Economics,’ ‘Business Finance,’ ‘Management,’ and ‘Business, econometrics, and finance’ as depicted Figure 1. Second, only the final versions of the peer-reviewed journal papers published in English were included. The filtering process returned a reduced dataset of 629 documents. To ensure data quality, additional manual screening was performed to confirm the relevance of these articles to AI and finance research.

Figure 1. Tree Map indicating initial query results.

2.4 Final Inclusion

Following comprehensive screening, 607 documents were finalized as highly relevant to the literature on AI in finance. Manual verification of the dataset was performed and each document provided substantial insights. These documents were included in the subsequent bibliometric analyses.

3.1 Publication trend

Since the mid-1900s, the concept of AI has experienced sizable growth. At the Dartmouth Conference in 1956, John McCarthy first used the word to describe the making of machinery that might carry out assignments that normally required human intelligence (McCarthy et al., 1956). Over time, AI has been expanded to encompass machine learning, deep learning, and neural networks, moving away from its initial focus on rule-based systems and symbolic reasoning (Mohamed et al., 2020; Russell & Norvig, 2016). AI’s increasing capacity to evaluate enormous datasets, adjust to new knowledge, and produce remarkably accurate forecast insights is shown in this progression.

AI’s transformative impact has been particularly profound in the financial sector, where its adoption has redefined traditional processes and decision-making paradigms. Financial institutions have leveraged AI for applications ranging from algorithmic trading and fraud detection to personalized customer services and risk assessment (Agrawal et al., 2018). For instance, AI powered algorithmic trading systems can process and analyze financial market data in real time, enabling high-frequency trades that maximize profitability (Baldauf & Mollner, 2020). Similarly, financial institutions can anticipate market patterns, identify possible weaknesses, and create plans to reduce losses by using AI-driven risk management systems (Cao, 2023; Jain, 2023).

AI is essential for improving customer engagement and personalizing experiences in marketing and consumer finance. AI-powered advanced recommendation engines examine consumer behavior and preferences in order to make personalized financial product recommendations, enhance client loyalty, and foster stronger relationships (Davenport & Ronanki, 2018). Furthermore, chatbots and virtual assistants provide real-time assistance with natural language processing (NLP) capabilities to revolutionize customer service in banking and financial services (Coglianese & Lehr, 2019; Shin, 2019). Recent bibliometric analyses of AI in finance literature stresses its interdisciplinary nature and rapid expansion as a field of study. For instance, studies have identified key trends in AI applications within financial services, highlighting an emphasis on predictive analytics, real-time decision making, and automation (Dhamija & Bag, 2020). Bibliometric research has also revealed the central role of collaboration among researchers from diverse disciplines, thus reflecting the field’s complex and integrative challenges. In this study we have further explored the ethical considerations and governance challenges accompanying AI’s widespread adoption in finance, emphasizing the importance of balancing innovation with regulatory oversight (Li & Zhang, 2024; Ridzuan et al., 2024; Taeihagh, 2021).

AI is a key force behind innovation in the financial industry because of its capacity to interpret unstructured data, identify hidden patterns, and make judgments on its own. Meanwhile, debates over fairness, transparency, and the possible replacement of human labor in finance-related positions have been sparked by its broad implementation (Cao, 2023; Pal et al., 2021). As AI develops further, its dual function as a tool to improve operational efficiency and ethical investigation highlights its great and complex effect on finance.

The status of scientific production for each year is shown in Figure 2 and Table 1. Our bibliometric research documents the scholarly investigation of AI in finance, identifying 607 papers published between 1993 and 2024, starting in the early 1990s. This area has shown consistent improvement, with a yearly growth rate of 3.49%. Throughout the first 25 years (1993–2017), academic output was relatively modest, with only 139 publications, averaging approximately 5.56 articles annually. However, this changed dramatically after 2018, with an exponential increase in the publication rates. From 2018 to 2023 alone, an average of 121.6 publications per year demonstrates the accelerating interest in AI applications within finance. The majority of research contributions result from this increase, suggesting that in recent years, the relationship between AI and finance has become a crucial topic of study.

Figure 2. Annual scientific production.

An average of 15.73 citations per document and 30,726 references showed a growing volume of research in the area. Furthermore, international collaboration is prominent in this domain, with 45.63% of publications involving authors from multiple countries highlighting the global relevance of AI in reshaping financial systems. Major applications explored in the literature include risk management, trading algorithms, customer relationship management, and decision-making tools. Scholarly interest increased sharply between 2019 and 2023, coinciding with advances in automation, big data analytics, and machine learning. The relatively young age of documents (2.98 years on average) suggests that this is an emergent field, characterized by fresh insights and cutting-edge research. Despite the recent proliferation of studies, the field is still in its formative stages, with single-authored contributions comprising only 46 articles, indicating the collaborative nature of this research domain. The expanding adoption of AI technologies by financial institutions and fintech companies, along with the emergence of data-driven solutions, is expected to sustain and even amplify this growth trajectory. The sharp increase in publications after 2018 highlights the transformative capability in this area. As the financial sector continues to embrace AI-driven innovations, further academic exploration is anticipated, promising to enhance our insight into the role of AI in revolutionizing financial systems and practices.

3.2 Productive Journals

The breakdown showed that 607 articles were published in 205 journals. We searched for journals with at least one publication and at least 25 citations to determine which were the most significant and active in furthering AI and finance research. 64 journals that satisfied the required limits were identified using this criterion ( Table 2). The journals were ranked according to the quantity of publications they published; the journal that published the most articles was ranked the highest. The rating was based on the overall number of citations earned by each publication, even if several journals had published the same number of articles. Our analysis indicates that Management Science, Journal of Banking & Finance, Technological Forecasting and Social Change, Finance analysis Letters, and Journal of Economic Dynamics & Control are the top five most productive publications for AI and finance research. The advancement of the literature in this field has been greatly aided by these journals. According to the ABDC journal rankings, these journals are notably ranked A or A*, demonstrating their significance and the significant influence of the research they publish. Table 2 provides a thorough summary of the journals, along with their citation and publication counts.

3.3 Top cited articles

To identify the most impactful contributions to this domain, we adopted a citation-based benchmarking approach, focusing on articles with at least 100 citations (Bhukya & Paul, 2023). This analysis identified ten seminal articles, which are detailed in Table 3. With an average of 36.89 citations annually (CPY), a research article (Das & Chen, 2007) published in Management Science has received 664 citations. The author (Altman et al., 1994), in the Journal of Banking and Finance, gathered 464 citations (14.97 CPY) by offering a comparative evaluation of neural networks and traditional methods for corporate distress diagnosis, a foundational contribution to predictive analytics in finance. Several other studies have also made profound contributions to the Journal of Economic Dynamics and Control (LeBaron et al., 1999) and received 386 citations (14.85 CPY) for their innovative exploration of time-series properties within artificial stock markets, bridging AI, and economic modelling. A nonparametric AI-driven method for pricing and hedging derivatives was presented (Hutchinson et al., 1994) in The Journal of Finance, which received 355 citations (11.45 CPY). More recently, the authors (Chod et al., 2020) highlighted the financial advantages of supply chain transparency and blockchain technology in Management Science, achieving 324 citations with an outstanding 64.80 CPY. The emerging significance of AI-driven decision making is further illustrated in the California Management Review (Huang et al., 2019) (200 citations, 33.33 CPY) and Finance Research Letters (Dowling & Lucey, 2023) (179 citations, 89.50 CPY). Despite being published only recently, the latter has demonstrated rapid scholarly engagement with AI in finance, emphasizing its critical role in shaping future research trajectories. The majority of these highly significant works were published in esteemed journals with ABDC classifications of A* or A, including Management Science, The Journal of Finance, and the Journal of Banking & Finance. Their large number of citations highlights a vital contribution to the growth of AI applications in the financial industry. This has laid the groundwork for future research. A thorough overview of these studies is given in Table 3, which also highlights the variety of approaches, research topics, and consequences that have influenced the academic environment of financial and AI studies. This emphasizes that AI is becoming a more significant factor in risk management, market analysis, and financial decision-making.

3.4 Top productive authors

Our analysis revealed that 1966 writers contributed to the field of study. Nineteen played a significant role in developing the research area. Our aim was to identify the authors with the most cumulative citations and the most prolific authors. In terms of total citations (tc = 528) from 13 articles, Andrew W. Lo of the Massachusetts Institute of Technology stood out as the top contributor, highlighting his noteworthy influence in the field. With eight publications and 84 citations overall, Zhaobo Zhu dominated the list of authors with the most articles. Additional noteworthy contributors include Shasha Zhou and Yuangao Chen, each with seven publications and 211 and 176 cumulative citations, respectively, underscoring their strong influence in advancing academic discourse. Furthermore, researchers such as Maoyong Cheng (three publications, tc = 169) and Xingyang Lv (three publications, tc = 152) have also demonstrated focused and impactful research output. These contributions collectively reflect the dynamic and interdisciplinary efforts that drive the growth of AI and finance research. Figure 3, Figure 4, and Table 4 provide insightful snapshots of these leading scholars, emphasizing their pivotal roles in shaping the scholarly landscape.

Figure 3. Author-network of top cited articles.

Figure 4. Productive and influential authors.

3.5 Co-occurrence Network Analysis

Figure 5 reveals the thematic structure of research on AI and Finance through a co-occurrence network in which distinct clusters of interconnected keywords are identified, each represented by a specific color. The largest cluster, highlighted in green, centers around “artificial intelligence” and includes closely linked terms like “machine learning,” “big data,” and “automation,” reflecting the widespread use of AI across diverse domains such as finance and technology. The red cluster emphasizes terms like “blockchain” and “cryptocurrency,” showcasing the amalgamation of AI with financial technologies, particularly in enhancing transparency, security, and digital innovation. The blue cluster focuses on “financial markets,” “risk assessment,” and “systemic analysis,” systemic analysis, demonstrating the use of AI in traditional financial research for predictive analytics and economic modelling. Additionally, smaller clusters, such as purple and orange clusters, capture niche areas such as behavioral analysis and decision-making systems. The multidisciplinary character of AI research is highlighted by these color-coded clusters, which integrate technological, financial, and economic development with computational advances. They also emphasized the crucial role that AI plays in tackling intricate problems in a variety of domains.

Figure 5. Co-occurrence network.

4.1 Theoretical Foundations of AI in Finance

The analysis reveals six major hypotheses that explain how AI transforms financial innovation and decision making. The Efficient Market Hypothesis (EMH) explores how AI enhances the accuracy of financial market predictions and identifies inefficiencies that contribute to improved market efficiency (Anshari et al., 2021; Buckley et al., 2021; Fama, 1970; Kurshan et al., 2021). Behavioral Finance Theories examine the influence of AI-driven sentiment analysis on investor psychology and decision-making, highlighting the psychological underpinnings of financial behavior (Chartier et al., 2021). Agency Theory highlights AI’s potential to mitigate agency conflicts by enhancing transparency, monitoring, and corporate governance mechanisms, thereby reducing information asymmetry (Jensen & Meckling, 1976; Kalkan, 2024). Decision theory examines how AI streamlines intricate financial decision-making procedures, especially in domains such as investment strategies and risk management (Cao, 2023; Tversky & Kahneman, 1986). Network Theory illustrates AI’s significant role in systemic risk analysis, particularly through the modelling of financial networks, enabling a deeper understanding of interconnected risks (Battiston et al., 2016). Finally, the Resource-Based View (RBV) positions AI as a strategic organizational resource that firms leverage to gain competitive advantages, innovate, and drive long-term growth (Barney, 1991; Kitsios & Kamariotou, 2021; Machireddy et al., 2021). These theories offer a solid basis for comprehending the complex effects of AI on the financial industry to improve predictive analytics, decision-making procedures, and strategic management of financial institutions.

4.2 Contextual Foundations of AI in Finance

The literature on AI applications in finance spans diverse contexts, each contributing distinct insights into the transformative potential and the associated challenges of AI-driven innovation. Among these, the financial market context highlights AI’s application in algorithmic trading, market sentiment analysis, and fraud detection, showcasing its ability to enhance efficiency and security in trading environments (Baldauf & Mollner, 2020; Cao, 2023). In Corporate Finance, AI is considered for financial forecasting, capital structure optimization, and strategic investment decisions, demonstrating its value in improving corporate decision-making processes (Machireddy et al., 2021). The Banking sector leverages AI-powered solutions for credit risk assessment, customer relationship management, and the automation of loan approval processes, thereby enhancing operational efficiency and customer satisfaction (Jain, 2023; Singh & Arora, 2023). Within Insurance, AI contributes to streamlining underwriting, managing claims, refining actuarial risk models, and fostering precision and cost efficiency (Erem Ceylan, 2022; Rana et al., 2022; Zia & Kalia, 2022). The rise of FinTech showcases role of AI in developing innovative tools, such as robo-advisors, chatbots, and blockchain-driven systems, revolutionizing customer interactions, and transaction processes (Kashyap, 2023). Finally, the Sustainability context explores AI’s contributions to environmental, social, and governance (ESG) investments, green finance, and carbon market analytics, highlighting its critical role in supporting sustainable economic development (Hemanand et al., 2022; Macchiavello & Siri, 2022). These six contexts collectively demonstrate the breadth of AI’s impact on finance, operational efficiency, strategic decision-making, and sustainability initiatives.

4.3 Characteristics Underneath AI in Finance

AI’s transformative impact of AI on finance is underpinned by its unique technical, functional, and strategic characteristics, which enable it to address complex challenges and unlock new opportunities. A key component of AI in finance is Data-Driven Insights, which use big data to identify trends, facilitate predictive analytics, and improve decision-making in a variety of financial fields (Go et al., 2020; Obschonka & Audretsch, 2020). Automation simplifies repetitive tasks, such as financial reporting, auditing, and trading, improves efficiency, and reduces operational burdens (Jain, 2023; Singh & Arora, 2023). The Scalability of AI allows its application to expand seamlessly across various financial sectors and global markets, demonstrating its adaptability and potential (Chen & Bellavitis, 2020; Jameaba, 2020). With Accuracy and Speed, AI delivers high-precision computations while accelerating complex financial processes, thereby ensuring timely and reliable outcomes (Cao, 2020, 2023; Jameaba, 2020). However, the use of AI also brings ethical challenges such as concerns about bias, equity, and transparency in AI-powered systems, which call for strong governance structures (Binns, 2018). AI also serves as a force for disruption, transforming traditional financial paradigms, fostering innovation, and driving new market dynamics (Kalkan, 2024). Together, these characteristics displays capacity of AI to enhance efficiency, reduce costs, and redefine strategies, making it a pivotal driver of financial innovation and evolution.

4.4 Methodological Foundations of AI in Finance

AI research within finance domain has adopted a range of methodological approaches to investigate its multifaceted implications and applications. Empirical Studies show that machine learning algorithms perform predictive modelling and financial analysis, offering data-driven insights into market behavior and trends (Cao, 2023). Experimental Designs focus on testing AI-based solutions for critical functions, such as fraud detection, trading strategies, and risk management, enabling practical evaluations of effectiveness of AI (Kashyap, 2023). Simulation Models create artificial markets and scenarios, thus providing a framework for assessing systemic risks and market dynamics under controlled conditions (Battiston et al., 2016). Qualitative Research delves into stakeholder perceptions and attitudes toward AI adoption in finance, shedding light on behavioral and organizational factors (Ridzuan et al., 2024). Mixed Methods integrate quantitative and qualitative approaches, offering a holistic perspective of impact and potential of AI (Kitsios & Kamariotou, 2021; Machireddy et al., 2021; Tang, 2021). Network Analysis employs AI-driven graph analytics to map financial systems and uncover relationships, thereby enhancing the systemic understanding (Jain, 2023; Singh & Arora, 2023). Additionally, Systematic Reviews synthesize existing literature to identify emerging trends and research gaps, guiding future inquiries (Baldauf & Mollner, 2020; Cao, 2023). Despite considerable advancements, there are still gaps in the use of cutting-edge approaches to better grasp the revolutionary role of AI in finance, such as agent-based modelling and neuroscience (Go et al., 2020).

5.1 Contextual Future Research Directions in AI and Finance

The industry has seen a dramatic change as a result of the rapid development of AI, which has opened new avenues for risk management, predictive analysis, and more effective decision-making. The use of AI in financial operations is still developing, and a number of important areas require further research. As technology advances and the financial landscape changes, more research is necessary to fully understand how AI propels innovation in markets and financial institutions. A promising path toward enhancing operational effectiveness and financial stability is provided by the AI’s capacity to handle enormous amounts of data and deliver insights into dynamic environments. With increasing influence of AI, particularly in areas such as algorithmic trading, fraud detection, and financial forecasting, the following future research questions (FRQs) can guide further investigation into the role of AI in finance.

FRQ1: How can AI-driven predictive models enhance decision-making in volatile financial markets, particularly during economic downturns or crises?

FRQ2: How does the use of AI to financial decision-making raise ethical questions, particularly with regard to accountability, transparency, and fairness?

FRQ3: What effects will AI deployment have on conventional risk management techniques and the part that human decision-makers play in these procedures in financial institutions?

FRQ4: How can AI innovations improve financial inclusion, particularly in underserved or emerging markets?

FRQ5: What are the longstanding effects of AI on investor behaviour and financial market dynamics, particularly regarding the automation of trading and advisory services?

FRQ6: Can AI play a role in enhancing sustainability in finance by supporting green finance initiatives and responsible investment practices?

5.2 Characteristics Future Research Directions in AI and Finance

AI in finance is driven not only by its technical capabilities but also by its potential to disrupt traditional financial models and improve efficiency. Future research can explore the interplay between AI technologies and behavioral factors that influence decision-making processes in finance. Key areas of interest include the following.

FRQ1: How does AI-driven sentiment analysis affect investor psychology and market movements, particularly in volatile or uncertain periods?

FRQ2: What role does AI play in reshaping financial advisory services, and how do consumers perceive the trustworthiness and effectiveness of AI-driven advice compared to traditional human advisors?

FRQ3: How can AI be leveraged to improve decision-making under uncertainty, particularly in complex financial markets where human cognition may be limited?

As AI technology continues to evolve, its interaction with human behavior and cognitive processes in finance remains a critical area for future research. Understanding how AI enhances or challenges human decision-making and the ethical implications of these changes is crucial for shaping the future of finance. Research in this area should consider the potential of AI to not only optimize financial outcomes but also its role in influencing consumer behavior and financial practices.

5.3 Methodological Future Research Directions

AI is used in a large range of financial products, from algorithmic trading to fraud detection, risk management, and financial forecasting. Quantitative methods, such as statistical analysis and machine learning models, are the mainstays of current research in this field. Model overfitting, data bias, and limited interpretability of sophisticated AI systems are some of the methodological problems that still exist. Future research can address these issues by implementing a greater variety of creative techniques to increase the precision, dependability, and interpretability of AI-driven financial models. Using text analytics and natural language processing (NLP) tools to examine vast volumes of financial data from many sources, such as business filings, social media, and financial news, is a possible path for methodological progress. Using named entity identification, topic modelling, and sentiment analysis, researchers can learn more about investor behavior, market mood, and patterns that affect financial decision making. These methods can provide more accurate predictions of market movements and risk factors while also addressing biases inherent in structured financial data.

Additionally, simulation methodologies such as agent-based modelling and fuzzy cognitive mapping offer promising approaches for understanding complex interactions in financial systems. These techniques allow for the modelling of dynamic financial environments where individual agents, such as investors or firms, interact under uncertain conditions. Research on how AI affects financial stability, risk propagation, and market behavior can be conducted using agent-based models. On the other hand, fuzzy cognitive mapping can offer a better understanding of systemic risk and role of AI in financial crises by visualizing and comprehending the interaction between AI-driven decisions and other macroeconomic factors. Lastly, qualitative approaches can be significant in comprehending the behavioral consequences of AI adoption, as AI continues to upend conventional banking practices. Stakeholder opinions about the use of AI in finance can be investigated using case studies, expert interviews, and interviews. The impact of AI on decision making, organizational behavior, and the future of financial advice services could be the main topics of discussion. Combining these qualitative insights with quantitative data can offer a holistic view of transformative potential of AI in finance.

6.1 Managerial Implications

Our research highlights the managerial implications for financial institutions and other stakeholders in the finance sector. Financial institutions should be mindful of AI’s potential to influence customer behavior in both beneficial and detrimental ways. While AI can drive efficiency and improve decision-making, its use in customer-facing applications must be approached with caution to avoid ethical concerns such as algorithmic bias or lack of transparency. Marketers, for example, should focus on promoting AI-based financial tools in ways that are transparent and aligned with consumers’ best interests, rather than relying solely on automation for decision-making.

6.2 Implications for Policymakers

Policymakers need to address the ethical and legal issues raised by AI in finance. It is imperative that governments and authorities put in place mechanisms that guarantee accountability, transparency, and justice in AI algorithms, given the widespread use of AI in decision-making. Oversight is required as AI systems in finance have become more sophisticated to prevent market manipulation, discrimination, and data privacy violations. Furthermore, regulators should focus on creating standards for AI literacy among consumers to help them understand the implications of AI-driven financial products/services.

6.3 Ethical Implications

AI in finance has several significant ethical implications. Although AI systems can improve decision making, they can also reinforce pre-existing biases, raising concerns about equality and fairness. Ethics pertaining to the independence of AI-driven judgments, openness in algorithmic procedures, and protection of consumer rights must be addressed as AI continues to have an impact on the financial industry. Researchers and practitioners should pay close attention to these issues, ensuring that AI adoption in finance does not disproportionately affect vulnerable populations or widen existing financial disparities.