Introduction

In the past decade, the explosion of Artificial Intelligence (AI) has significantly contributed to various aspects of life, addressing numerous complex challenges (Tripathi et al. 2024). The development of AI has become a phenomenon, with scholars consistently researching AI-related topics over the past few decades (Verma et al. 2021). One of the most widely cited definitions of AI is “a system’s ability to interpret external data correctly, to learn from such data, and to use those learnings to achieve specific goals and tasks through flexible adaptation” (Haenlein and Kaplan 2019).

In marketing and brand management, many terms equivalent to AI are used. This is because, AI encompasses numerous functions and related concepts, as it is often defined broadly to include various types of computers that can directly perform or support tasks that previously required human emotions or cognition, using software and algorithms (Haenlein and Kaplan 2019; Kumar et al. 2019). Consequently, research in the fields of marketing and branding often applies terms such as machine learning, chatbot, deep learning, neural network, automation, knowledge engineering, big data, interactive agent, expert system, text mining, data mining, soft computing, fuzzy logic, IoT (internet of things), robotics, intelligent retrieval, computer system, natural language processing, as taxonomies or thesaurus terms of AI (e.g., Martínez-López and Casillas 2013; Kumar et al. 2019; Salminen et al. 2019; Varsha et al. 2021; Verma et al. 2021; Vlačić et al. 2021).

AI has revolutionized branding, changing consumer and brand interactions (Mustak et al. 2021). Defined as creating and maintaining a brand (Musaiqer and Hamdan 2023), branding now relies on AI to influence consumer choices (Musaiqer and Hamdan 2023). AI enables real-time data analysis, enhances customer behaviour insights, drives loyalty, and improves experiences (Tjepkema 2019). It identifies consumer desires and guides organisational development (Verma et al. 2021). AI’s impact includes algorithmic decision-making, automation, machine learning, and social media analytics (Singh et al. 2019). These innovations foster personalized experiences and stronger brand-consumer relationships (Musaiqer and Hamdan 2023)

AI’s impact on branding has garnered significant research attention, particularly recently (Mustak et al. 2021). Understanding AI’s rapid evolution is crucial for technology-driven marketing (Varsha et al. 2021), leading to exponential growth in related research (Varsha et al. 2021; Mustak et al. 2021; Verma et al. 2021). Foundational theories addressing AI’s role in branding include brand personality, equity, identity, positioning, and resonance theories (Varsha et al. 2021). However, research suggests the field is still nascent (Vlačić et al. 2021), prompting calls for deeper exploration of AI’s impact on marketing and branding (Musaiqer and Hamdan 2023; Varsha et al. 2021; (Davenport et al. 2020).

Despite the crucial role of AI in branding and the increasing interest from scholars, efforts to synthesize knowledge on this topic remain limited, uneven, and unclear (Stone et al. 2020; Varsha et al. 2021). Due to its emerging nature, there appears to be a lack of a clear framework for understanding how AI relates to brand research (Varsha et al. 2021). Therefore, there is a growing need for comprehensive literature reviews on the impact of AI on branding. To date, to the best of our knowledge, only one study by Varsha et al. (2021) has examined 117 articles in the Scopus database from 1982 to 2019. Subsequently, there seems to be no further effort to synthesize knowledge on this topic.

Meanwhile, with the proliferation of Chat GPT and similar new AI technologies in late 2022, the number of publications on the relationship between AI and branding has exponentially increased in recent years (Mustak et al. 2021). There is an urgent need for comprehensive studies on this important topic. Varsha et al. (2021) have significantly contributed by emphasizing clusters within this theme. However, the number of documents reviewed by Varsha et al. (2021) is still limited, focusing only on the period from 1982 to 2019, and additional thesaurus terms or taxonomy of AI should be added to ensure more comprehensive coverage. Furthermore, there has been no study analyzing some important aspects of the topic such as growth trajectory, or the most influential authors. A novel review could provide more comprehensive and updated insights into the impact of AI on branding.

The purpose of this study is to review research on the impact of AI on branding using bibliometric analysis, incorporating scientific mapping and descriptive analysis. The research questions (RQs) of this paper include:

RQ1: How has the volume and growth trajectory of research on the impact of AI on branding evolved?

RQ2: How is the geographical distribution and collaboration between countries in research on the impact of AI on branding?

RQ3: Who are the most productive and influential authors in research on the impact of AI on branding?

RQ4: What are the most influential documents in research on the impact of AI on branding?

RQ5: What are the main schools of thought, and directions for future research on the impact of AI on branding?

Methodology

Bibliometric analysis is a scientific research method proposed by Pritchard (1969). With this method, researchers can combine scientific mapping and descriptive statistics (Hallinger and Kovačević 2022). Scholars can use co-authorship, co-occurrence, co-citation, and bibliographic coupling, thus enabling the synthesis of knowledge on a topic quantitatively and objectively (Zhao et al. 2019).

Identification of sources

The Scopus database was chosen because of its broader scope, which allows for advanced search capabilities to filter data conveniently, aiding researchers in analyzing and managing data more effectively (Verma et al. 2021, Hallinger and Nguyen 2020). Scopus uses a consistent rule for selecting documents for its index and includes a broader range of materials for evaluating social science research compared to Web of Science (Hallinger and Chatpinyakoop 2019).

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al. 2009) were utilised for document search and screening. The PRISMA guidelines emphasise comprehensive search strategies, detailed data extraction, and structured reporting (Kumar et al. 2023), which enhance reliability - key objectives of bibliometric studies. The application of PRISMA significantly improves the quality of reporting in systematic reviews and meta-analyses. (Sewell et al. 2023). In this study, the steps of PRISMA are outlined in Fig. 1.

Fig. 1
figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) in the identification and screening of sources.

Full size image

In the initial step, to identify documents, the initial search string included terms such as “artificial intelligence” (or thesaurus terms) and brand* searched in the search fields: Title, Abstract, and Keywords. The thesaurus terms for AI were inherited from Martínez-López and Casillas (2013), Varsha et al. (2021), Verma et al. (2021), and Vlačić et al. (2021). As a result, we obtained 8,025 initial documents.

In the second step of screening, we applied several inclusion criteria as follows: (1) Documents published until 2023; (2) Articles published in journals as they represent validated knowledge (Podsakoff et al. 2005); (3) The field of business, management, and accounting; (4) Articles in English. Given the relatively broad scope of this paper, the initial search yielded a high number of documents (8025). Therefore, we applied multiple criteria as outlined. While the research topic is interdisciplinary, its primary focus is on branding, leading us to restrict the dataset to the fields of business, management, and accounting. This is a common approach to exclude irrelevant literature, especially when the research topic is broad and overlaps with multiple domains (e.g. Verma et al. 2021, Vlačić et al. 2021). Narrowing the scope of a study, such as selecting a specific field or limiting keywords, enhances data management and avoids unnecessary complexity in interdisciplinary research (Donthu, Kumar, Mukherjee et al. 2021). When necessary, restricting the dataset allows for a sharper focus on the most meaningful aspects of the research, ensuring that the analysis results are both reliable and highly practical (Zupic and Čater 2015).

The data, after screening, comprised 708 documents, which were downloaded in both Excel and text formats. It is important to note that for interdisciplinary topics, the exclusion rate of documents retrieved from Scopus search results tends to be high (Hallinger and Chatpinyakoop 2019).

In the third and fourth steps, the two authors read the titles, abstracts, and full texts when necessary to identify relevant documents. Ultimately, we retained 592 documents for analysis. The excluded documents were studies that were largely unrelated to the topic under consideration.

Data analysis

To analyse the data, we utilised Excel and VOSviewer (Van Eck and Waltman 2010). Excel was employed for descriptive statistics regarding growth trajectory, volume, the most productive countries, the most influential authors, documents, and sources. VOSviewer was used for co-authorship analysis and bibliographic coupling techniques. Co-authorship analysis was employed to map knowledge structures reflecting the geographical distribution and international collaboration in the research topic (Koseoglu 2016). Bibliographic coupling analyses were utilised to visually represent the main schools of thought through a scientific map based on the analysis of situations where two documents cited the same third work (Koseoglu 2016).

Findings and Discussions

Volume and growth trajectory

Figure 2 illustrates the volume and growth trajectory of 592 articles indexed in Scopus from 1982 to 2023. Overall, there has been a rapid increase in research on this topic over the past four decades.

Fig. 2
figure 2

Number of articles and growth trajectory.

Full size image

During the period from 1982 to 2013, the number of articles focusing on the impact of AI on branding was quite limited, accounting for only 9.1% (with 54 articles over 22 years, averaging 2.45 studies per year). The first publication related to this topic was relatively early, in 1982. Over the next two decades, this topic did not receive significant attention from scholars, with the number of articles remaining in single digits.

The period from 2014 to 2018 witnessed an increase in articles on this topic. In 2014, the number of publications reached double digits for the first time and continued to steadily increase each year. However, the number of documents during this period was still limited, accounting for 14.4% (with 85 articles over five years, averaging 17 studies per year).

The period from 2019 to 2023 saw a significant increase in scholarly interest in the impact of AI on branding. In 2019, there was a remarkable surge in research output, with 57 articles, more than twice the previous year. Subsequently, the number of studies on this topic increased rapidly each year, peaking in 2023 with 117 articles. The total number of studies during this period accounted for 76.5% (with 453 articles over six years, averaging 75.5 studies per year).

Such discoveries regarding the historical development of the topic can be explained by several reasons. Academic interest in AI in brand management began in the 1980s, with the emergence of perspectives suggesting that AI tools could support decision-making (Collins and Mauritson 1987) and effectively forecast sales (Steinberg and Plank 1987), followed by studies focusing on Early Expert-Systems and robotics (e.g. Gill 1995). After more than two decades of relative quiet, the increased interest of scholars and practitioners in brand management may be related to three main factors: the breakthrough of AI techniques and supporting technology tools, the evolution of big data, and the availability of computational power (Overgoor et al. 2019; Bock et al. 2020).

Countries and international collaborations

Regarding the contributing countries, Fig. 3 displays the global distribution of documents on the topic across 76 countries spanning various continents, highlighting widespread interest in AI’s impact on branding. Node size in Fig. 3 reflects the prominence of this topic among researchers worldwide, with significant attention from scholars in the United States, the United Kingdom, India, and China. Notably, research from Austria, Ireland, Singapore, and Germany emerged earlier, followed by contributions from the United States, the United Kingdom, and more recently, India, South Korea, and China.

Fig. 3: International distribution and collaborations in research on the impact of AI on branding.
figure 3

(threshold: 1 document per country, 76 countries).

Full size image

Table 1 provides detailed information on the number and percentage of articles from the top 10 most productive countries in this field. Consistent with Fig. 3, the results show that the United States is the most prolific, with 175 documents, accounting for nearly 20% of the global research output. The United States also has the highest citation count. Following the United States is the United Kingdom with 73 documents and India with 67 documents. Both Fig. 3 and Table 1 indicate that the countries most interested in this topic are primarily the United States, various European nations, several Asian countries, and Australia, while other continents contribute less significantly.

Table 1 The most productive countries.
Full size table

Geographical analysis of AI adoption in branding reveals national and regional trends shaped by cultural and economic factors. In the United States, the global leader in AI application research in branding, the significant productivity of scholars can be attributed to advanced technological infrastructure, strong consumer markets, and high levels of innovation in digital marketing. This can also be traced back to the earliest findings on the role of AI in business originating from the United States, such as Allen (1982), Collins and Mauritson (1987), Steinberg and Plank (1987), and Gill (1995). The United States is also home to major tech companies that lead the way in deploying AI solutions for branding and customer interaction. Additionally, the country boasts strong academic institutions and abundant research funding, facilitating the development of AI across various fields, including branding.

In the United Kingdom, the adoption of AI in branding is likely influenced by a strong digital economy and a well-developed advertising and marketing industry. Research in the United Kingdom often focuses on integrating AI technology to enhance customer experience, personalize brands, and optimize marketing strategies. The tradition of data-driven decision-making and the growing focus on ethical AI in the United Kingdom likely also contribute to the country’s strong position in this field.

India’s high output in this area can be attributed to the rapidly growing information technology and tech industry, coupled with a massive consumer market, creating favorable conditions for the application of AI in branding. Companies in India increasingly leverage AI to serve a diverse, tech-savvy population, particularly in fields like e-commerce, mobile applications, and digital advertising. The Indian government also promotes digital initiatives and AI projects, accelerating research and development in this area.

The rise of China in AI branding research can be directly linked to the government’s strategic focus on AI in its national development plans. China’s rapidly growing digital economy and the widespread use of AI technologies in consumer services and retail have created significant opportunities for applying AI in branding. Major tech companies like Alibaba and Tencent have made substantial investments in AI to transform customer interactions and build brand loyalty. Furthermore, China’s large population and high mobile device usage create an ideal environment for applying AI in branding and customer engagement.

Our findings are somewhat consistent with previous related studies. For example, Varsha et al. (2021) reviewed the literature on the impact of AI on branding from 1982 to 2019 and identified the United States, Germany, and France as the top three countries publishing the most research. Mustak et al. (2021), in examining AI in marketing, noted that the bulk of research primarily comes from three regions: the United States, Central and Southern Europe, and East Asia.

Regarding international collaborations, the number and thickness of lines in Fig. 3 indicate significant links between countries. Particularly noteworthy is the research partnership between the United States and the United Kingdom, demonstrating the highest total link strength of 13. This is followed by collaborations between the United Kingdom and India, with a total link strength of 10, and between the United States and Canada, with a total link strength of 9. Other significant collaborations include those between China and the United States, China and the United Kingdom, and the United States and Australia, each with a total link strength of 8.

The collaboration trends shown in Fig. 3 suggest that cultural and economic dynamics may have influenced these partnerships. For example, the strong link between the United States and the United Kingdom can be attributed to similarities in language, culture, and shared research priorities, while the collaboration between the United Kingdom and India reflects the historical relationship and India’s growing role in global technological innovation. The increase in AI branding research in East Asia, particularly from South Korea and China, reflects the region’s investment in AI technology and digital consumer behaviour.

Influential authors

Our data reveal that there are 1581 authors with at least one study on the impact of AI on branding, and 1,469 of these authors have at least one citation. This indicates that the field has garnered significant attention from researchers worldwide.

Table 2 lists the top 10 most prolific authors and the most frequently cited authors in the Scopus database within this field. Leading the list of the most productive authors are Loureiro Sandra Maria Correia from Portugal and Fronzetti Colladon Andrea from Italy, each with seven documents. They are followed by Paulo Rita, Sérgio Moro, João Guerreiro (all from Portugal), and Silvia Ranfagni from Italy, along with Damianos P. Sakas from Greece, each with five documents. Notably, five out of the top ten authors are from Portugal. This finding aligns with Table 1 and Fig. 3, which show that Portugal ranks among the top 10 most productive countries in this research area.

Table 2 The most productive and influential authors.
Full size table

A deeper look into this reveals a few reasons for these results. Most of these prolific authors are affiliated with the University Institute of Lisbon, which hosts eight diverse research units. Among them, the Business Research Unit (BRU-IUL) stands out as a key unit, receiving substantial funding from the Foundation for Science and Technology (FCT) and other sources. This unit boasts numerous researchers and offers doctoral programs in data analysis, marketing, economics, and management. The institution has developed excellent research groups in marketing, management, and data analysis, creating an environment conducive to interdisciplinary research advancements.

Leading the list of the most cited authors are Goh Khim-Yong and Heng Cheng-Suang from Singapore, and Lin Zhijie from China, each with 1,038 citations. This prominence is understandable as these three scholars co-authored a highly influential paper (Goh et al. 2013), which is a typical study exploring social media brand communities and consumer behaviour, as well as the effects of user- and marketer-generated content. Following them are Raffaele Filieri from France with 781 citations, and Philipp Rauschnabel from Germany with 778 citations.

Notably, no author appears on both the lists of the most productive and the most cited authors. This indicates that while the group of researchers from Portugal and some others have made significant efforts to publish numerous papers, none of their works have achieved substantial influence in the field. Conversely, some authors from Singapore and other countries have not published many papers, yet their research has had a significant impact. Despite the United States being considered the most productive country in terms of the number of publications, it does not have any authors who are among the most influential in terms of citation count.

Influential documents

Table 3 presents the most cited papers in the Scopus database. Leading the list is the study by Goh et al. (2013) with 1038 citations. Following that, Dwivedi et al. (2021), and Lee et al. (2018) occupy the second and third positions with 738 and 527 citations, respectively. Overall, most of these studies primarily focus on the impacts of social media. For example, they examine the relative impact of user-generated versus marketer-generated content on social media brand communities (Goh et al. 2013), consumer engagement on Facebook (Lee et al. 2018), digital marketing (Dwivedi et al. 2021), visual content and social media engagement (Li and Xie 2020), measuring social media influence (Arora et al. 2019; Kiss and Bichler 2008), and real-time co-creation (Buhalis and Sinarta 2019). Additionally, some studies explore the impact of online conversations or e-service chatbots (Chung et al. 2020; Tirunillai and Tellis 2014), indicating that participation in social media brand communities leads to increased spending.

Table 3 The most influential documents.
Full size table

Two notable studies are Goh et al. (2013), Dwivedi et al. (2021). Goh et al. (2013) demonstrated the effects of user-generated content (UGC) and marketer-generated content (MGC) on purchasing behaviour through embedded information, providing evidence that UGC has a stronger impact than MGC on purchasing behaviour, thus playing a positive role for brands. The opinion paper of Dwivedi et al. (2021) has garnered the second-highest total citations, within just three years, averaging about 246 citations per year. Dwivedi et al. (2021) synthesized insights from several leading experts in digital marketing and social media communication, thereby providing detailed information and key insights into this topic, such as AI, mobile marketing and advertising, electronic word-of-mouth, ethical issues in marketing, augmented reality, and digital content management.

Main schools of thought

Bibliographic coupling analysis was employed to explore the main schools of thought on the impacts of AI in branding. Using a threshold of at least one citation per document and a total of 390 documents, we identified six main schools of thought. Figure 4 illustrates these clusters. Table 4 presents detailed information on the number of documents and the most cited studies in each cluster.

Fig. 4: Main schools of thought on impacts of AI on branding.
figure 4

(threshold 1 citation, 390 documents).

Full size image
Table 4 The main schools of thought of research on impact of AI on branding.
Full size table

The first school of thought: the integration of AI in branding through Chatbots, voice assistants, and AI influencers

This school of thought (in red colour) tends to focus on topics such as the effectiveness of Chatbot e-services in retail, the evolution of virtual assistants in marketing, building trust with AI voice assistants, proposing conceptual frameworks for robotics adoption in customer service, and examining consumer engagement with brands through AI influencers. For instance, Chung et al. (2020) find that Chatbots provide interactive and engaging brand or customer service encounters, thus supporting their adoption for virtual assistance. Mustak et al. (2021) identify dominant AI-related research themes in marketing, including the roles of Chatbots and virtual assistants. Pitardi and Marriott (2021) show that social presence and cognition are crucial for trust, illustrating a dynamic between privacy and trust in user interactions. Xiao and Kumar (2021) highlight the impact of robotics on service quality and customer engagement, moderated by firm nature, service characteristics, and brand positioning. Trivedi (2019) reveals that quality dimensions of information systems significantly impact customer experience and brand love, moderated by perceived risk, offering strategic directions for enhancing consumer-brand relationships.

The second school of thought: The intersection of social media and AI in brand management

Scholars in this school of thought (in green colour) are interested in investigating various aspects of the impact of social and digital media on consumer behaviour, brand engagement levels, and marketing strategies. They explore topics such as the transformation of consumer interactions through social media, measuring the impact of influencers, enhancing smart destinations through big data analytics, and differentiating destination branding concepts. The findings highlight the opportunities and challenges posed by digital technology and social media, as well as the effectiveness of data-driven approaches in improving branding strategies. For instance, Dwivedi et al. (2021) highlight opportunities and challenges. Buhalis and Sinarta (2019) conclude that real-time consumer intelligence and data-driven approaches revolutionize service co-creation by enabling dynamic, personalized consumer experiences. Marine-Roig and Clavé (2015) indicate key metrics like engagement and sentiment are crucial for determining influencers, with an ensemble model achieving the highest accuracy in predicting influencer index. Moro et al. (2016) show that automated web content mining effectively extracts destination brand identity and image from online sources, providing valuable insights for branding strategies.

The third school of thought: the influence of UGC and MGC on consumer behaviour and brand development

This school of thought (in blue colour) investigates various aspects such as the influence of UGC and MGC on consumer purchase behaviour, the association between online advertising content and consumer engagement, the extraction of consumer satisfaction dimensions from UGC, and the utilization of big data for brand management insights. Additionally, they delve into the detection and response strategies for negative electronic word of mouth (eWOM) and the role of visual content in driving customer engagement on social media platforms. For instance, Goh et al. (2013) find that UGC has a stronger impact than MGC, particularly in undirected communication modes. Lee et al. (2018) highlight the effectiveness of combining brand personality-related content with informative content to improve engagement. Tirunillai and Tellis (2014) investigate that subjective dimensions dominate horizontally differentiated markets, while objective dimensions are more prevalent in vertically differentiated markets.

The fourth school of thought: leveraging advanced analytical approaches in branding through neural networks, sentiment analysis, and AI

The common aim of these studies (in yellow colour) is to explore innovative methods and technologies for understanding consumer behaviour, assessing brand performance, and managing brand equity. Key findings include the application of neural networks to model consumer choices, the development of measures like the semantic brand score to assess brand importance, the use of sentiment analysis and AI to distinguish between fake and real news content and the investigation of factors influencing the adoption of new products by low-income consumers in emerging markets. For example, Bentz and Merunka (2000) show neural networks capture non-linear preferences and aid logit models. Roy et al. (2019) find smart services impact brand equity and word-of-mouth. Roberts et al. (2014) highlight big data and digital communication’s influence on brand management. Colladon (2018) introduces the Semantic Brand Score for measuring brand importance. Pournarakis et al. (2017) combine topic and sentiment analysis for brand performance insights.

The fifth school of thought: navigating consumer experience, insights, and branding strategies in the AI age

The scholars in this school of thought (in purple colour) aim to investigate various aspects of consumer behaviour, perception, and experience in the context of technological advancements, marketing strategies, and brand dynamics. They provide deep insights into how factors such as the Internet of Things, viral marketing, online advertising, sensory perception, and brand personality influence consumer decision-making and brand engagement. For instance, Hoffman and Novak (2018) identify four types of consumer experiences in the IoT era and advocate for a nonhuman-centric approach. Kiss and Bichler (2008) find central customers enhance viral marketing, but the best centrality measure depends on network topology and diffusion. Liu and Mattila (2017) reveal that advertising appeal effectiveness for Airbnb varies with an individual’s sense of power. Chan and Tung (2019) note that robotic service enhances sensory and intellectual experiences, with varied impact on affective and overall brand experiences across hotel segments. Wu et al. (2017) find that a friend-like interaction style in IoT environments enhances brand warmth, competence, and attachment. Chen et al. (2015) show that brand personality traits are encoded in brain regions associated with reasoning, imagery, and affective processing.

The sixth school of thought: Crafting consumer engagement strategies and ensuring brand authenticity in the AI era

These studies (in light blue colour) focus on the impact of online customer engagement dimensions, emotions, and consumer-generated media stimuli on brand engagement and authenticity. Additionally, the scholars in this cluster tend to focus on the role of machine learning and AI integration in enhancing consumer engagement and brand value. For instance, Bilro et al. (2019) find that cognitive processing and hedonic experience significantly impact online customer reviews, stressing the importance of meeting consumer expectations. Loureiro et al. (2019) reveal that website quality, pleasure, and arousal boost consumer-brand engagement, emphasizing emotionally appealing content. Aluri et al. (2019) show that machine learning excels at identifying customers who value specific promotions, enhancing engagement and loyalty. Ballestar et al. (2019) use machine learning to personalize financial incentives, optimizing digital marketing returns. Schivinski (2021) identifies five consumer engagement subtypes through machine learning, offering tailored strategies for brand-related social media behaviour.

Future Research Directions

The following future research directions are proposed based on the intellectual structure of the field discussed in earlier sections, a systematic synthesis of prior studies, the pressing research questions identified in those studies, the suggested research avenues from scholars (Donthu, Kumar, Pandey, et al. 2021; Nascimento and Loureiro 2024).

Firstly, the ethical and responsible use of AI in branding is a crucial future research direction. As AI becomes more prevalent in branding, organisations must ensure its ethical and responsible use, as advocated by current research (Musaiqer and Hamdan 2023). This includes ensuring that AI is not used to harm customers or create biases, and that customer data is collected and used transparently and honestly (Musaiqer and Hamdan 2023). Future research could focus on developing standardized guidelines to ensure trust, transparency, and authenticity when using AI. For instance, scholars might examine how businesses can prevent social discrimination and ensure that AI is applied inclusively without widening the wealth gap (Mustak et al. 2021). As AI advances, it has the potential to significantly alter the nature and size of the workforce in the marketing industry and create digital divides between regions with and without access to modern technology (Huang and Rust 2018). Future studies need to focus on how to implement AI inclusively and equitably, avoiding social discrimination and preventing the widening of wealth disparities. This requires marketers not only to understand theoretical principles but also to apply AI ethically to enhance brand equity without violating ethical standards (Mustak et al. 2021).

Secondly, the impact of AI on a company’s stakeholders in the context of sustainable brand development is a key area for future research. Current studies primarily focus on how companies use AI to enhance marketing functions and customer relationships through information, knowledge, and technology. However, there is a noticeable gap in research addressing how to manage and apply AI across the entire marketing network and its role concerning other stakeholders (Davenport et al. 2020). Today, sustainable branding is emerging as a trend and a competitive advantage. Strengthening relationships with stakeholders is increasingly recognized as a measure of corporate sustainability, yet knowledge in this area remains relatively limited (Winit et al. 2023; Donthu et al. 2021; Nascimento and Loureiro, 2024). Comprehensive research can help companies better understand how to leverage AI to optimize branding activities from both internal and external perspectives, aiming for sustainability.

Thirdly, hyper-personalization to enhance brand image, trust, and loyalty is a promising direction for future research. To deliver highly tailored experiences, AI will continue to harness vast amounts of customer data (Musaiqer and Hamdan 2023). As noted, AI offers profound insights into consumer behaviour, providing recommendations for actions that can secure and maintain customer loyalty, influence their future behaviours, and holistically evaluate their experiences (Tjepkema 2019). AI techniques are instrumental in identifying customer preferences and guiding the future development of businesses (Verma et al. 2021). Therefore, future research can focus on how companies use AI to deeply understand customers, enhance brand experiences, and meet their needs without infringing on their privacy. This research direction should ideally be integrated with the first, as enhancing personalization must be aligned with the ethical collection and use of individual customer data.

Fourthly, enhancing the application of advanced AI technologies in brand development, such as optimizing voice and visual search, augmented and virtual reality, semantic understanding is a vital future research direction. As more people use voice and visual search to find goods and services, AI will become increasingly important in optimizing content for these methods (Musaiqer and Hamdan 2023). Future studies can focus on developing advanced algorithms to enhance brand image and equity by improving the accuracy and relevance of search results. Brand managers need to find ways to leverage AI-powered smart speakers and wearable devices to optimize customer experiences (Mustak et al. 2021). Additionally, the trend towards rich virtual reality experiences will be further enhanced by AI (Musaiqer and Hamdan 2023). Future research can explore how AI can help customers better interact with brands through augmented and virtual reality technologies. The use of AI to gain deeper insights into customers through semantic understanding and machine learning is an emerging research direction that requires further exploration due to the explosion of big data and recent advancements in AI. Algorithms inspired by psychology and neuroscience are increasingly improving the ability to predict consumer behaviour. Future studies could investigate how to use deep learning models to detect complex emotions, thereby enhancing AI-based brand recognition.

Fifthly, autonomous marketing plays a crucial role in brand development and the involvement of marketers within that context. Data analysis, content creation, and decision-making are among the marketing aspects being automated by AI to support brand building and development (Musaiqer and Hamdan, 2023). This aids brands in reducing operational costs and enhancing their adaptability to customer needs. Future research could focus on optimizing human intervention and AI automation to maximize brand development efficiency. According to Haenlein and Kaplan (2019), using AI without understanding its mechanisms can lead to adverse consequences. Managers may not grasp how AI systems they rely on function, leading to strategic and tactical errors if unaware of how machine learning algorithms operate or their profound influence on individual decisions and actions (De Bruyn et al. 2020). This raises a critical question about the timing and extent of automating decisions instead of leaving them to humans, and whether there exists an ideal balance between automated tasks and those carried out by humans (Mustak et al. 2021). According to Mustak et al. (2021), pertinent questions to consider include: What roles should humans undertake when designing AI-based marketing systems? When should decisions be automated rather than human-controlled? Is there an ideal balance between tasks automated by AI or performed by humans, considering the nature and extent of tasks involved?

Conclusions and Limitations

The overarching aim of this study is to examine research on the impact of AI on branding indexed by Scopus from 1982 to 2023. Utilizing bibliometric indices, this research tracks the proliferation of documents, revealing the geographical distribution and international collaboration in research, leading authors, most influential documents, and the main schools of thought of this field. Specifically, the research findings indicate a significant increase in the number of studies on the role of AI in branding over the past four decades, peaking in 2023. This topic has garnered attention from scholars worldwide, with major contributions coming from the United States, the United Kingdom, India, and China. Importantly, the research has revealed six main schools of thought, and five directions for future research in the field.

This study has certain limitations. Due to the characteristics of bibliometric methods and the length limitations of articles, it is difficult for authors to fully present the detailed content analysis results of studies on this topic. Additionally, not all parts of the bibliometric analysis, such as co-citation analysis and keyword co-occurrence, could be included in this article due to word limit constraints. Future research could combine quantitative bibliometric analysis with qualitative content analysis to provide deeper insights. Scholars in the future could also focus on other analytical techniques of bibliometric methods to complement the gaps in this paper. Besides, due to the interdisciplinary nature of the topic and the vast amount of literature, this study only focuses on scientific articles indexed on Scopus, published in English, and limited to the field of business, management, and accounting. Therefore, it may overlook valuable studies beyond the scope of document screening. Future research should explore this topic in various fields, languages, types of documents, databases, and extend the publication time frame.