A Survey on Digital Trust: Towards a Validated Definition

Abstract

Digital trust is increasingly crucial for successful interactions in modern digital environments. However, the existing literature lacks a unified definition and a comprehensive understanding of its core factors. This study addresses these gaps by conducting a systematic literature review to explore and synthesise existing definitions of digital trust and identify the fundamental factors that shape it. A total of 86 relevant sources were analysed, revealing that digital trust is typically conceptualised as confidence in people, processes, and technology aimed at ensuring a secure digital environment, with data protection and privacy playing critical roles. Through thematic analysis, “openness” emerged as an additional factor complementing previously established elements of the integrative model of organisational trust, such as ability, benevolence, and integrity. Based on 42 definitions, we developed a new holistic definition of digital trust. The authors evaluated its content validity, confirming its alignment with the essential factors shaping digital trust’s essence. The findings highlight the multidimensional nature of digital trust and offer an operationalised framework for future measurement and application.

1. Introduction

As technology becomes increasingly central to our lives and businesses, understanding and fostering digital trust (DT) is crucial for maximising the potential of digital technologies and maintaining healthy relationships with them [1]. Online environments provide more significant uncertainty and perceived risk than traditional settings, making trust a key factor in reducing fear and enabling engagement [2]. According to Alpcan, Levi, and Savas [3], there is a fundamental difference between trust relationships within online and offline environments due to differences in trust establishment settings; namely, in traditional trust formation, personal interaction plays an important role. At the same time, the concept of DT appears prominently in professional literature as a new approach to assessing the trustworthiness of digital technologies [4,5,6,7].

Research on DT has grown significantly since 2016, with studies focusing on its definition, formation process, and impact on various sectors, but it is still developing in scientific literature [2]. Trust should be considered contextual and varies across different domains [8,9]. Dimitrakos connects it with the competence of the other party (trustee) in behaving dependably within a given context and relative to a specific task [10].

Digital interactions experienced tremendous growth during the last decade; from 3.5 billion Internet users in 2016 [11], it transformed to 5.56 billion in 2025 [12]. This motivates exploring new dimensions that bring the digital environment into the trust concept [13]. The authors of various papers stated that there is no single definition of DT [2,14,15]. This leads to a limited understanding of this concept and its dimensions, preventing organisations from establishing DT [15]. Sometimes alternative terms are used, such as “cyber-trust” or “online-trust” [14]. Nevertheless, no single definition or clear understanding of alternative terms’ differentiation exists.

Simultaneously, researchers noticed a significant difference between “traditional” trust and DT. While DT includes the features of regular trust, as it is built on fundamental trust foundations, primarily social and psychological factors [16,17], it also has several unique characteristics due to technological or digital factors that influence it [3,16,17]. The need to explore DT resulted in several systematic literature review (SLR) research papers on DT topics, three of which were included in the scope of this review.

The first study of Tunkevichus and Rebiazina concentrated exclusively on consumer DT and identified the main trends and research directions in this field. The second systematic literature review, conducted by Pietrzak and Takala [2], had a broader scope, focusing on the concept of DT. While this work provided valuable insights into DT at the time, several limitations highlighted the need for a new systematic literature review rather than merely a replication with newly published sources:

• The authors used only one database (Web of Science, WoS) for their research, significantly limiting the number of sources;
• The inclusion and exclusion criteria were not described in their paper [2].

The third identified SLR by Sharma et al. [18], concentrated on adopting digital trust in blockchain-based supply-chain management, analysing the current state and providing future research directions. Thus, previous research has not solved the issue of various DT interpretations and the lack of consensus on its characteristics despite a highlighted strong need [19]. Therefore, the aim of this research is to conduct comprehensive research on the essence of the DT concept, including its definition and key factors. Further, considering the absence of a unified definition, the authors proposed their own holistic definition and tested it with the content validation method. It is important to acknowledge that the research devoted to trust in the digital environment became topical at least a decade before the DT concept appeared. For example, the “e-trust” concept focused on consumer trust in online government started to appear in the late 1990s [20]. However, this paper does not aim to look into the formation of various concepts related to trust in the digital environment and systematically distinguish the nuances between them.

2. Materials and Methods

This study comprised two phases. First, an SLR was conducted to gather a comprehensive theoretical foundation for the existing definition of DT and to analyse its factors. Second, the content validation method [21] was employed to demonstrate how the proposed definition of the DT concept aligns with its key factors.

The SLR was conducted per PRISMA 2020 methodology [22], which provides a standardised protocol to transparently report how the sources were reviewed and included in the final scope. The papers were searched in two of the world’s leading bibliographic scientific databases [23]—Scopus [24] and WoS [25]. The search, using the keyword “digital trust”, took place on 15 September 2024 and was restricted to (1) titles, abstracts, and keywords; (2) scientific articles and conference papers; and (3) the English language. Conference papers were included in the review, given that the concept of DT is still emerging.

A total of 156 records were identified in Scopus [24] and 84 in WoS [25]. After removing duplicates, 170 papers remained. All abstracts were subsequently retrieved and reviewed. The inclusion criterion was a mention of the DT concept in the text as a concept discussed within the paper. In turn, authors excluded papers that discussed DT primarily from a computer-science or engineering perspective; for example, from technical frameworks or algorithms for trust (e.g., cryptography, smart contracts, trusted execution environments, secure protocols) perspective, or detailed engineering solutions to trust/privacy/security problems, since the aim of this review was to look at DT as applicable for the social science. Following the initial abstract screening, 118 sources qualified for full-text review.

The authors could not access the full texts of 12 reports. Therefore, 106 papers were read, and 20 reports were excluded based on the previously stated criteria. As a result, 86 research papers—27 conference proceedings and 59 scientific articles—were included in the theoretical analysis of the DT concept. The PRISMA 2020 flow diagram is shown in Figure 1 below.

3. Systematic Literature Review Results

3.1. Bibliographical Analysis

The bibliographical analysis revealed that the first publication included in the analysis appeared in 2010 as conference proceedings. Interest in DT began to increase significantly in 2019, with four papers published, peaking in 2022 with 21 publications—five times more than in 2019. With 16 papers retrieved by September 2024, there is an expectation that the total number of publications in 2024 will equal or exceed that of 2022 if the publication rate remains consistent in the coming months. The papers included in the review, sorted by publication year and type, are illustrated in Figure 2.

These facts motivate an investigation of the external factors that could contribute to the significant growth of research devoted to DT. From a regulatory perspective, mid 2018 marked the enforcement of the European Union’s General Data Protection Regulation, prompting the companies to prioritise privacy and data protection [26]. Simultaneously, the California Consumer Privacy Act was signed in June 2018 in California, United States, granting users new rights over personal data and imposing transparency duties on businesses [27]. The Cambridge Analytica scandal, which involved the collection of personal data from millions of Facebook users without their explicit consent by Cambridge Analytica, a political consulting firm, became public in 2018, significantly heightening public awareness of data-privacy issues and the potential for misuse of personal information by online platforms and political entities [28].

Further on, COVID-19 accelerated digitalisation in early 2020, increasing reliance and dependency on digital services more than ever before, as commerce, payments, learning, and communication rapidly shifted to the digital environment [29]. This might be why Edelman’s 2021 Trust Barometer global online survey of 31,000 respondents, conducted in October and November 2020, showed a significant 7% drop in public trust in technology [30].

While trust in the technology industry began to rebound the following year (end of 2021), growing by 4% [31], Edelman’s online survey conducted in 2023 showed that trust in the technological sector (76%) coexists with near-distrust in innovative technology (artificial intelligence, machine learning, natural language processing, and generative AI). Moreover, 59% of respondents stated that government regulators “lack adequate understanding of emerging technologies to regulate them effectively” [32] (p. 16).

Finally, the number of data breaches and cybersecurity incidents continues to grow. The annual number of data compromises in the United States grew from 1108, with 310 million individuals affected in 2020, to 3158, with over 1.35 billion individuals affected in 2024 [33]. Therefore, the accelerated digitalisation post COVID-19, combined with fluctuating public confidence and growing data-security concerns, could raise the actuality of such studies.

The keyword analysis revealed 328 keywords. A co-occurrence “full-counting” keyword analysis using VOSviewer [34] identified 11 keywords that appeared three or more times. The relatively small number of recurring keywords indicates a diverse focus among the analysed articles, reflecting the multifaceted nature of the DT concept.

Aside from “digital trust” (n = 39) and “trust” (n = 17), the most frequently mentioned keywords were “blockchain” (n = 15), “digitalisation” (n = 6), “digital transformation” (n = 6), and “sharing economy” (n = 6). These trends indicate a strong link between digital transformation and blockchain technology, highlighting contemporary research subjects focused on the shift from traditional to digital environments, which enable new business models.

The clustering analysis identified four clusters, as depicted in Figure 3:

• Blue: devoted to governance;
• Red: focused on digital transformation and digitalisation, closely connecting them with industrial or organisational changes;
• Green: related to the technological application of DT;
• Yellow: concerned with socio-economic aspects.

These findings highlight that the blockchain appears as the main technology used to build trust in digital settings. They also illustrate how trust is essential for adopting new digital processes and economic models.

3.2. Existing Definitions of Digital Trust

The research revealed 42 various definitions of DT in academia. Some authors provided their own definitions, whereas others referred to one of a few definitions from other sources. Notably, within the wide range of the identified definitions, only five were cited several times within the scope of the analysed articles. These observations suggest that no single proven point of view exists within academia, and authors have divergent views on the DT concept.

Nevertheless, based on the analysis, the most popular definition of DT is “the confidence of stakeholders in the competence of actors, technologies and processes for establishing reliable and secure business networks” [35] (p. 3), which was included in the analysed papers and quoted three times [36,37,38], i.e., used four times in total.

The complete list of identified definitions, classified into six thematic groups by applying content analysis, is presented in Appendix A. Some definitions encompass several themes but have been allocated to the most thematically appropriate group. The thematic distribution of the definitions is summarised in

Table 1. DT definitions distribution by the thematic group ¹.

Group nr.	Main Group’s Theme
Group 1 (n = 12)	Confidence in people, processes and technology to create a secure digital environment
Group 2 (n = 8)	Trust in data protection, privacy, security, and ethical information handling
Group 3 (n = 8)	Trust in the technical capability and reliability of technology
Group 4 (n = 7)	Trust as a mental state involving cognitive, emotional, and experiential factors
Group 5 (n = 4)	Trust as an enabler in digital relationships, interactions, and business networks
Group 6 (n = 3)	Trust based on ethical behaviour and cultural principles

¹ Source: created by authors.

below.

Further, each thematic group is described in detail, with examples of common traits and definitions.

Group 1: Confidence in people, processes, and technology to create a secure digital environment. This is the most prevalent definition type, represented by 12 examples. The common traits of these definitions are:

• A focus on collective confidence in people, processes and technology, i.e., “holistic trust”.
• An emphasis on having a secure and reliable digital environment.
• Includes trust from various stakeholders such as users, customers, individuals, partners, and society.

For example, Das et al. [39] (p. 360) define DT as follows: “in the digital context, trust refers to users’ confidence and reliability in the systems, services, and organisations they interact with. Users must trust that their personal information will be handled responsibly, that their privacy will be respected, and that the systems they rely on are secure from malicious actors.

Group 2: Trust in data protection, privacy, security, and ethical information handling. This thematic group is also well represented, with eight examples. These definitions are focused chiefly on information security and management:

• Assurance that personal and sensitive data are protected from unauthorised access and breaches.
• Expectation in the organisations to handle data responsibly and ethically, respecting user privacy.
• Meeting stakeholders’ expectations regarding data-handling practices.

Notably, this was the type of definition proposed by Pietrzak and Takala [2] (p. 64) as the result of their systematic research on DT: “Digital trust is assumed to be the measure of confidence that workers, consumers/buyers, partners, and other stakeholders have in an organisation’s ability to protect data and the privacy of individuals”.

Group 3: Trust in the technical capability and reliability of technology. Eight definitions fall under this thematic group, whose common patterns may be described as:

• Confidence in technology’s ability to execute tasks and transactions accurately and securely.
• Trust in the consistent performance and punctuality of digital systems, platforms, and services.
• Use specific technologies (e.g., blockchain, AI) to establish or enhance trust.

An example is: “The consumer’s belief that the service is technically capable of ensuring the successful execution of the transaction” [40], quoted in [14] (p. 433).

In this regard, it is essential to note that some researchers portray blockchain, which can be defined as “a digital distributed ledger that secures and links the digital records called “blocks” using cryptographic techniques” [41] (p. 2) as DT, and there is a scientific discussion on this topic. For instance, Shin states DT is “a kind of user heuristics in blockchain” [42] (p. 2), meaning that when blockchain technology is used, users tend to assume that the digital solution is trustworthy. Some papers also name blockchain as a default guarantee, a substitute for DT [43,44]. While this technology can partially substitute trust between actors in certain collaborative, cooperative, contract-based activities, that does not fully substitute trust. For example, it can record transactions and automatically execute smart contracts but cannot address ethical dilemmas or resolve conflicts that emerge during the contract’s formation [45]. DT cannot be achieved solely through technology; it requires building trust in institutions and processes [46].

Group 4: Trust as a mental state that encompasses cognitive, emotional, and experiential factors. This group comprises seven definitions that are less digital/technical than previous groups. Its common patterns include:

• Trust involves rational evaluations (competence, integrity) and emotional responses (feelings, beliefs).
• Trust is built upon past experiences and evidence of behaviour, shaping future expectations.
• Trust affects the willingness to take risks and make decisions, even with limited verification.

For example, Shin [42] (p. 2) defines “digital trust as cognitive heuristics constitute information processing methods to make decisions more quickly and with less effort than more complex methods, and thus they reduce cognitive load during security assessment”, and Hochstein et al. [47] (p. 1) “the willingness to rely on digitally presented information when there are limited means of verification”.

Group 5: Trust as an enabler in digital relationships, interactions, and business networks. This theme, represented by four definitions, highlights the following common aspects:

• Trust facilitates interactions in environments involving human actors and technological elements.
• Trust is crucial for customer acquisition, retention, and the formation of reliable business networks and partnerships.
• Technology can be an enabler that either enhances or impedes trust in digital relationships.

An example of this thematic group is: “Digital trust represents the acquisition and retention of customers and shareholder value via providing confidence in the digital services with digital channels” [48], quoted in [49] (p. 6).

Group 6: Trust based on ethical behaviour and cultural principles. This is the least represented thematic group, appearing in three definitions. Its main characteristics are:

• Trust is based on digital partners’ integrity, benevolence, and predictability.
• Trust encompasses adherence to both written and unspoken commitments to avert harm.
• It emphasises privacy, security, protection, and data management as part of an organisation’s culture.

For instance: “the digital trust category is also a general term to describe behavioural and cultural principles, including privacy, security, protection and data management” [50], quoted in [51] (p. 544).

According to these thematic groups, there is no unified approach to defining DT. Consequently, it is essential to clarify this concept, as the lack of familiarity hinders progress in scientific research and causes organisations to encounter challenges in measuring, managing, and enhancing their level of DT.

Identifying the key factors that contribute to DT is necessary to construct a comprehensive definition that reflects the multidimensional and holistic nature of the concept.

3.3. Key Characteristics of Digital Trust

The importance of a shared understanding of the DT concept and its characteristics is evident from previous attempts to identify the characteristics of DT using a SLR methodology. The recent publication dates of these studies further reinforce this point. The need for a consensus on DT and its characteristics among key stakeholders worldwide was also highlighted by the World Economic Forum’s Centre for Cybersecurity [19].

However, both previous studies primarily focused on blockchain as the main DT provider and drew on blockchain-related literature to extract DT characteristics.

Rychkova and Ghriba [52] reviewed blockchain literature to identify 21 technology-neutral trustworthiness requirements, emphasising the importance of digital, technological, and social trust in organisational decision making. Based on the trustworthiness factors Mayer, Davis, and Schoorman [9] mentioned, the requirements were split into three categories: ability, benevolence, and integrity.

Sharma, Agrawal, and Manupati [18] examined the adoption of DT in blockchain-based supply-chain management and listed its characteristics. They identified five key characteristics of DT essential for sustainable blockchain-based supply chains: transparency, cybersecurity, data protection, accountability, reliability and provenance, and regulatory compliance. This study offered more high-level characteristics, as some factors could subsume additional factors identified by Rychkova and Ghriba [52].

The present SLR focuses exclusively on DT but was not restricted to the blockchain aspect. A total of 36 analysed papers included the characteristics of DT, yet none appeared in a systematic review by Rychkova and Ghriba [52]. From these papers, 24 DT factors were identified and organised into four categories and eight sub-categories.

Some authors used “trust” or “trustworthiness” as DT factors, which does not clarify the specific elements of digital trustworthiness. In reviewing various factors, it was noted that authors discussed them using different levels of granularity. Some were using exact “ability”, “benevolence”, and “integrity” [53]; some were naming categories, such as “security” and “responsible use” [54], whereas some were naming very detailed factors, such as “data privacy” [38] or “decentralisation” [55].

Because this research aims to identify universally applicable DT factors, all identified factors were arranged into three layers:

• Category, based on the initial trustworthiness factors introduced by Mayer, Davis and Schoorman [9];
• Sub-category, representing the foundational elements of the category in the context of DT; and
• Consolidated DT factors, as numerous researchers employed various elements to convey a single DT factor.

Trustworthiness requirements are linked to social science factors (ability, benevolence, and integrity) to bridge social and technological domains, following the approach of Rychkova and Ghriba [52]. As stated by Mayer et al. [9], all three categories—ability, benevolence, and integrity—are distinct yet interrelated. According to Mayer et al.’s definition, ability is “that group of skills, competencies, and characteristics that enable a party to influence within some specific domain” [9] (p. 717). Considering DT’s context, “ability” here includes:

• A system’s, service’s, and service provider’s competence and performance;
• Security and control factors, i.e., internal technical capabilities required to operate reliably;
• Secure system architecture factors representing internal design choices to enhance systems or services’ capabilities and perform reliably.

Such an interpretation may be discussable. For instance, it differs from Rychkova and Ghriba’s [52] classification, which places security and control under Category 3, “integrity”. Nevertheless, such a position goes against the initial definition of “integrity”, which means “that the trustee adheres to a set of principles that the trustor finds acceptable” [9] (p. 719). Implemented technical controls are characteristics of products or services rather than inherent values or correspondence to externally defined principles. While governance and compliance frameworks may prescribe certain standards, implementing security and control remains a matter of technical capability.

Table 2. DT factors included in Category 1: ability ¹.

Category 1: Ability	Source
1. Competence and reliable performance (cumulative n = 19)
1.1. Competence (n = 2)	[52,56]
1.2. Performance (n = 2)	[52,57]
1.3. System’s reliability (incl. safety and quality) (n = 8)	[16,46,54,55,58,59,60,61]
1.4. Operational stability (incl. availability, resilience and business continuity) (n = 7)	[16,49,52,54,57,59,61]
2. Security and control (cumulative n = 25)
2.1. Identity and access management (inc. authentication, authorisation, safe credentials and identifiability) (n = 6)	[2,49,52,62,63,64]
2.2. Confidentiality (incl. data protection and control over own data) (n = 9)	[18,51,52,54,62,64,65,66,67]
2.3. Integrity of data (incl. immutability) (n = 10)	[39,45,52,54,55,57,61,62,68,69]
3. Secure system architecture (cumulative n = 8)	[52,68]
3.1. Automation (n = 2)
3.2. Decentralisation (incl. consensus algorithms, permissionless access, censorship resistance and cryptographic functions/protocols) (n = 5)	[45,52,54,55,68]
3.3. Interoperability (n = 1)	[52]

¹ Source: created by authors.

below provides a comprehensive overview of the DT factors included in Category 1.

The second category, benevolence, is “the extent to which a trustee is believed to want to do good to the trustor, aside from an egocentric profit motive” [9] (p. 718). Mayer et al. [9] describe benevolence as the perception of a positive orientation of the trustee toward the trustor. The category can be considered a blend of cognitive and emotional trust, as it arises from the trustor’s subjective perception, the information available, and his experiences. In other words, this category is considered as the customer’s rating of the service or product based on his previous experience and information available to him. Among the identified factors, this category may be classified into:

• Relational credibility includes elements like goodwill, the reliability of the supplier or service provider, and their reputation.
• User experience and the level of support received—how satisfied the trustor is with the service, their prior experiences, and the convenience of using the product or service.

In this category, it is essential to distinguish between the service’s technical reliability and the user’s perception of whether they can depend on the service provider.

Table 3. DT factors included in Category 2: benevolence ¹.

Category 2: Benevolence	Source
1. Relational credibility (cumulative n = 10)
1.1. Goodwill (incl. benevolence, fairness and honesty) (n = 3)	[53,70,71]
1.2. Reliability (predictability of or confidence in service provider) (n = 3)	[2,53,61]
1.3. Reputation (incl. credibility) (n = 4)	[2,69,72,73]
2. User experience and support (cumulative n = 12)
2.1. Satisfaction of customer service and support (incl. prompt resolution of the issues) (n = 6)	[39,56,57,58,60,61]
2.2. Positive past experience (n = 3)	[59,72,73]
2.3. Usability (n = 3)	[52,71,72]

¹ Source: created by authors.

below presents the complete list of the DT factors in the benevolence category.

The earlier definition of integrity pertains to whether an organisation operates within the agreed-upon values, ethics, and rules. From the identified factors, this category can be divided into:

• The organisation’s core ethical principles include integrity, which guides its adherence to data ethics.
• Governance and compliance encompass accountability for the organisation’s actions, adherence to regulations, and the implementation of standards, including industry best practices and other frameworks.

Table 4. DT factors included in Category 3: integrity ¹.

Category 3: Integrity	Source
1. Core ethical principles (cumulative n = 14)
1.1. Integrity (the principle of the organisation) (n = 1)	[52]
1.2. Data ethics (incl. responsible use and privacy) (n = 12)	[38,42,46,49,52,54,59,62,65,66,73,74]
1.3. Sustainability (n = 1)	[36]
2. Governance and compliance (cumulative n = 13)
2.1. Accountability (n = 6)	[18,39,45,49,52,75]
2.2. Compliance with regulations (n = 6)	[17,18,52,67,74,75]
2.3. Application of standards (n = 1)	[72]

¹ Source: created by authors.

, with all DT factors within Category 3, is depicted below.

Although nearly all identified factors can be categorised into three main categories: 1. The system or organisation’s competence/capacity (ability); 2. The positive intent of organisations/service providers towards users (benevolence); and 3. The organisation’s adherence to acceptable standards and principles (integrity); one particular set of DT factors does not entirely conform to these three pillars. While Rychkova and Ghriba [52] placed “auditability”, “transparency”, and “traceability” under “Integrity”, defining it as “the moral quality of being sincere, honest, and consistent in one’s behaviour; capacity and willingness to adhere to some rules/principles” (p. 82), we argue that the methods for verifying trustworthiness through external sources and assurances go beyond an organisation’s internal adherence to ethical and governance principles and should be assessed as a separate component of DT. The body of reviewed sources, for example [39,55,70,76,77], looked into transparency and traceability from the perspective of blockchain technology, as essential for DT. This is also a key difference comparing with the offline settings, when trust may be ensured through personal interaction.

Several papers treat transparency, auditability, or traceability as a separate dimension of DT. For example, Jelovac et al. adopted a five-dimensional DT model that included transparency and integrity as separate dimensions [69]. Likewise, Das et al. treat them as separate metrics for trustworthiness evaluation [39]. In the modern world, trust is becoming demonstrable and publicly verifiable, and the segregation of this dimension highlights a need for its assessment and measurement. Therefore, this concept can be described as “openness” and could be defined as the extent to which an organisation or service makes its operations, data flows, and decision-making processes verifiable by external parties. Based on our analysis, this category encompasses three components:

• Transparency signifies open and clear information about how the service functions.
• Auditability pertains to an external party’s capacity to examine the service or its provider and provide independent verification or certification. In this context, the experiences of other users may also be emphasised, whether through feedback, reviews, or word of mouth.
• Traceability refers to the ability to track processes and data flows within the service.

As this category targets a single dimension and includes only three factors, no sub-category has been introduced, and cumulative referencing has been calculated for the entire category. The complete list of references for this category is presented in

Table 5. DT factors included in Category 4: openness ¹.

Category 4: Openness (Cumulative n = 23)	Source
1. Transparency (n = 11)	[18,39,45,49,52,55,59,62,70,75,78]
2. Auditability (incl. certification and independent verification) (n = 6)	[36,45,52,59,62,73]
3. Traceability (incl. data provenance) (n = 6)	[2,18,52,62,63,78]

¹ Source: created by authors.

below.

The overall framework of the factors included in the DT concept is depicted in Figure 4 below.

Following the preceding thematical analysis of the existing DT definitions, the authors propose to a holistic definition based on the two most widespread groups that highlight the DT’s nature of providing confidence in people, processes, and technology for secure digital environment creation, as well as a focus on data protection, privacy, and ethical data handling. Simultaneously, the proposed definition includes all components of the DT key factors: “Digital trust is the confidence that an entity or digital environment consistently demonstrates competence and reliability, maintains robust security and a secure system architecture, fosters fair and honest practices, provides a positive user experience, complies with governance and regulatory standards, and ensures auditability and traceability of data and operations”.

4. Digital Trust Definition Validation

Clear conceptual definitions are very important for scientific progress and further theoretical development [79]. Even more important is the creation of a formal definition using precise language to prevent unclarities and repetitive reasoning [80]. Although one of the 42 DT definitions appeared four times in the sample, it is difficult to label it as dominant. Furthermore, this definition was included in the first thematic group of definitions, and the authors of this study argue that it reflects the holistic nature of the concept.

4.1. Definition Validation Methodology

Considering the various existing definitions while proposing a new comprehensive one, there is a noticeable need for content validation, defined as “the methodological process of gauging the degree to which scale items adequately sample the universe of content associated with a construct” [81] (p. 1243). Although DT can be viewed as an interdisciplinary concept that merges technology with psychology, the methodology offered by Colquitt et al. [81] (commonly applied in industrial/organisational psychology) for conducting definitional correspondence (to analyse the degree to which a scale’s items or factors composing the concept correspond to its definition) and definitional distinctiveness (to analyse the extent to which a scale’s items or factors composing the concept correspond more to the given definition than to other related constructs) was considered relevant and was applied in this study.

The questionnaire for content validation of the definition was created using Hinkin and Tracey’s approach [21]. Since it was a self-guided online survey designed to determine the degree to which the definition corresponded to the key factors that compose it, the response scale offered by Colquitt et al. [81] in the detailed instructions (illustrated in Figure 5 below) was deemed unsuitable for the objectives of this study.

The respondents were informed that the research aims to validate the newly developed definition to ensure it effectively captures the essence of the DT concept. They were also informed that they would be invited to rate four definitions: three existing ones and one newly developed, in a random order. They evaluated how well each survey item, i.e., each characteristic, aligns with the four provided definitions.

The authors, fluent in both English and Latvian, first created the questionnaire in English before translating it into Latvian. This was essential for effective distribution in Latvia, the authors’ place of allocation, as not all potential respondents had the necessary level of English proficiency. Furthermore, the translation was validated by two native Latvian speakers who were fluent in English to ensure that both the definitions and the factors matched the original intent of the English versions.

Subsequently, the initial version of the questionnaire was piloted with five respondents to gather feedback and assess its validity. Following the pilot, the wording of the scale items was simplified, as the respondents found it difficult to grasp the essence of each factor and ascertain whether it was reflected in the provided definition. Additionally, based on the feedback, the Likert scale was reduced from 7 points to 5 points, as the respondents were uncertain about the fundamental difference between the “extremely bad/good” ratings and the “very bad/good” ratings, in line with Hinkin and Tracey’s original approach [21]. The final evaluation scale ranged from 1 to 5, where

• 1 means that the characteristic is not presented in the definition at all.
• 2 means that the characteristic is more absent than presented.
• 3 means that the characteristic is partially presented.
• 4 means that the characteristic is more presented than absent.
• 5 means that the characteristic is fully presented in the definition.

Seven DT subcategories were operationalised to measure the validity of DT definitions. For the fourth category, known as the openness category, no subcategory was formulated; however, two of its selected factors—auditability and traceability—were operationalised. These two factors were chosen because they both essentially provide some of the transparency promised by a digitally trustworthy organisation.

The operationalised descriptions of the validation factors, which are listed in

Table 6. Description of measurable items ¹.

Code	Category	Operationalised Description
Category 1: Ability
F1	1. Competence and performance	The entity demonstrates competence and reliably delivers services or products.
F2	2. Security and control	The entity ensures secure access, protects data, and maintains its integrity.
F3	3. Secure system architecture	The entity employs a secure architecture and processes to ensure efficiency, control, and integration.
Category 2: Benevolence
F4	1. Relational credibility	The entity demonstrates fairness, honesty, and reliability, building a strong reputation.
F5	2. User experience and support	The entity ensures a positive user experience and provides responsive support.
Category 3: Integrity
F6	1. Core ethical principles	The entity demonstrates integrity, protects data privacy, and adopts sustainable practices.
F7	2. Governance and compliance	The entity ensures governance and compliance in accordance with regulations and standards.
Category 4: Openness
F8	1. Auditability	The entity ensures compliance through certifications, audits, and supervision.
F9	2. Traceability	The entity ensures traceability by providing information about data origins.

¹ Source: created by authors.

below, were based on the factor descriptions proposed by Rychkova and Ghriba [52]. To operationalise the first three categories (F1–F7), the authors summarised the descriptions of the factors included in the sub-category, as provided by Rychkova and Ghriba [52]. The fourth category factors (F8, F9) were created directly using the descriptions from Rychkova and Ghriba [52].

Four definitions were provided for the respondents’ judgement: the first, developed by the authors for validation; the second, the most popular existing definition, to determine whether the proposed one better captures the concept’s essence; and the third and fourth, pertaining to the DT concepts of “cyber-trust” and “online trust”.

Definition No. 1: “Digital trust is the confidence that an entity or digital environment consistently demonstrates competence and reliability, maintains robust security and a secure system architecture, fosters fair and honest practices, provides a positive user experience, complies with governance and regulatory standards, and ensures auditability and traceability of data and operations” (proposed by the authors).

Definition No. 2: “Digital trust is the confidence of stakeholders in the competence of actors, technologies and processes for establishing reliable and secure business networks” ([35] p. 3), the most popular definition within the analysed sample.

Definition No. 3: “The user’s confidence in the predictability of the ‘behaviour’ of software and hardware systems (digital technologies), their reliability, which is manifested in the willingness to delegate several tasks to various software and hardware systems” [82] (cited in [14] p. 433). This defines the related concept of “cyber-trust”.

Definition No. 4: “Consumers’ perception of a web site’s usefulness, security, privacy, reputation, quality, and e-vendors’ willingness to customise” [83] (p. 433). The definition of the related concept is “online trust”.

Although Colquitt et al., (2019) [81] based on an analysis of 112 studies, stated that naïve respondents can be used when applying the methodology, doctoral students are considered the best judges for definitional correspondence and distinctiveness. This is because they look beyond the simple meaning of definitions due to their expertise and experience. Therefore, the authors tried to obtain input from this group of respondents and distributed the questionnaire among doctoral students through Facebook [84] and LinkedIn [85] PhD student groups and personal networks. The respondents were not compensated for their participation in the survey. However, to encourage completion, participants could provide an email address to enter a prize draw of three gift cards of the winner’s choice.

To ensure the necessary number of responses, and considering that, according to the methodology, the judges may also be naïve, the questionnaire was distributed among students at all levels at the university where the authors are affiliated—BA School of Business and Finance—and among professionals connected with digital topics through their networks. A total of 282 potential respondents opened the survey, 171 began to fill it out, and only 99 submitted their responses. Since Colquitt et al. [81] stated that a minimum sample size of 100 respondents would yield sufficiently precise results, we conclude that the current sample is adequate for this study.

4.2. Definition Validation Study Results

Statistical calculations in this study were conducted using Jamovi v.2.3.28 [86] software. A total of 47 (47.5%) respondents completed the questionnaire in English, whereas 52 (52.5%) selected the Latvian language for their responses. There was an option to select ages above 65, but none of the respondents chose it. The most represented age group was 35 to 44 (n = 34, 34.3%), and most respondents were under 44 (n = 83, 83.8%). From the educational perspective, the most represented group was those with a master’s degree (n = 43, 43.4%). Most respondents (n = 79, 79.8%) had obtained at least a college degree. The full demographic statistics are presented in

Table 7. Demographic statistics ¹.

Language	N	%	Age	N	%	Education	N	%
English	47	47.5%	18–24	21	21.2%	High-school diploma or equivalent	20	20.2%
Latvian	52	52.5%	25–34	28	28.3%	College	10	10.1%
			35–44	34	34.3%	Bachelor’s degree	26	26.3%
			45–54	13	13.3%	Master’s degree	43	43.4%
			55–64	3	3.0%

¹ Source: created by authors.

below.

Furthermore, the authors’ proposed definition (definition no. 1) was analysed according to the index of definitional correspondence, known as htc (for Hinkin–Tracey correspondence), calculated using the following formula:

htc = average definitional correspondence rating/a

The analysis of definition number 1 (D1) ratings confirmed that the judgments are not normally distributed, as anticipated in this study, based on the Shapiro–Wilk test; p < 0.001 in all instances. The standard deviation (SD) indicated some variation in the opinions; nonetheless, they remain largely consistent.

Following Hinkin and Tracey [21], a minimum value of 4.2 for htc provides a confirmation of content adequacy. According to Colquitt et al., (2019) [81] the htc value between 0.87 and 0.90 indicates strong content adequacy, while values from 0.84 to 0.86 suggest a moderate correlation that remains sufficient. Overall, the analysis reveals that the provided definition is moderately or strongly aligned with the key factors of the concept, enabling the authors to assert that it captures the essence of DT. While the judges noted that certain measurable items (such as competence and performance, user experience and support, and auditability) were not as strongly represented in the definition compared to others (like security and control or governance and compliance), their htc value still indicates that the definition adequately encompasses all assessed key factors. The calculations are presented in

Table 8. Statistical calculations of definition no. 1 ¹.

Code	Category	SD	p	Rating 2	htc 3
F1	Competence and performance	1.173	<0.001	4.18	0.84
F2	Security and control	0.969	<0.001	4.41	0.88
F3	Secure system architecture	0.959	<0.001	4.28	0.86
F4	Relational credibility	0.956	<0.001	4.32	0.86
F5	User experience and support	1.082	<0.001	4.18	0.84
F6	Core ethical principles	1.036	<0.001	4.26	0.85
F7	Governance and compliance	0.873	<0.001	4.48	0.90
F8	Auditability	1.152	<0.001	4.17	0.83
F9	Traceability	0.894	<0.001	4.34	0.87
Overall				4.29	0.86

¹ Source: created by authors. ² Rating: the average rating of each characteristic alignment with the provided definition. ³ htc: the index of definitional correspondence, which suggests a moderate correlation with a value from 0.84 to 0.86.

below.

The statistical analysis of the second (D2), third (D3), and fourth (D4) definitions confirmed that they were also not normally distributed, which was anticipated in this research. The SDs for the ratings of these definitions were higher than for the first definition, indicating that the ratings displayed greater variability from the judges. The calculations are provided in

Table 9. Statistical calculations of definitions no. 2, 3 and 4 ¹.

Code	Category	D2 SD	D2 p	D3 SD	D3 p	D4 SD	D4 p
F1	Competence and performance	1.003	<0.001	1.035	<0.001	1.082	<0.001
F2	Security and control	0.955	<0.001	1.218	<0.001	1.204	<0.001
F3	Secure system architecture	1.082	<0.001	1.051	<0.001	1.109	<0.001
F4	Relational credibility	1.069	<0.001	1.190	<0.001	1.063	<0.001
F5	User experience and support	1.094	<0.001	1.123	<0.001	1.071	<0.001
F6	Core ethical principles	1.041	<0.001	1.211	<0.001	1.023	<0.001
F7	Governance and compliance	1.114	<0.001	1.228	<0.001	1.161	<0.001
F8	Auditability	1.264	<0.001	1.229	<0.001	1.159	<0.001
F9	Traceability	1.161	<0.001	1.123	<0.001	1.218	<0.001

¹ Source: created by authors.

below.

Afterwards, the results of the similar (D2) and alternative (D3, D4) definitions were analysed using the htd or Hinkin–Tracey distinctiveness index, which was calculated using the formula below:

htd = Average of all (Intended Correspondence Rating − Orbiting Correspondence Rating)/(a − 1), where a is the number of anchors

According to Colquitt et al. [81], the htd value from 0.18 to 0.26 is moderate, whereas the value from 0.27 to 0.34 is strong. Following the calculations of the factors/measurable items’ ratings presented in

Table 10. Ratings ¹ and htd ² calculations for definitions no. 2, 3, and 4 ³.

Code	Category	D1 Rating	D2 Rating	D2 htd	D3 Rating	D3 htd	D4 Rating	D4 htd
F1	Competence and performance	4.18	3.65	0.13	3.36	0.21	3.53	0.16
F2	Security and control	4.41	3.62	0.20	3.31	0.28	3.59	0.21
F3	Secure system architecture	4.28	3.47	0.20	3.43	0.21	3.35	0.23
F4	Relational credibility	4.32	3.41	0.23	3.15	0.29	3.48	0.21
F5	User experience and support	4.18	3.13	0.26	3.27	0.23	3.71	0.12
F6	Core ethical principles	4.26	3.42	0.21	3.06	0.30	3.35	0.23
F7	Governance and compliance	4.48	3.23	0.31	3.11	0.34	3.17	0.33
F8	Auditability	4.17	3.21	0.24	3.00	0.29	3.06	0.28
F9	Traceability	4.34	3.17	0.29	3.06	0.32	2.92	0.36
Overall		4.29	3.37	0.23	3.20	0.27	3.35	0.24

¹ Rating: the average rating of each characteristic alignment with the provided definition. ² htd: distinctiveness index, which suggests the proposed definition’s distinctiveness level from other analysed definitions (moderate: 0.18–0.26; strong: 0.27–0.34). ³ Source: created by authors.

below, definition no. 2—the most popular definition of DT—is moderately distinctive from the newly proposed definition. Thus, we may conclude that the proposed definition (D1) better represents the conceptual essence of DT than the existing definition.

In turn, definition no. 3 (D3)—the definition of cyber trust—exhibits strong distinctiveness, whereas definition no. 4 (D4)—the definition of online trust—demonstrates moderate distinctiveness compared to the proposed D1. From this, we can conclude that the proposed D1 represents the DT concept, rather than the orbiting ones.

5. Discussion

The thematic analysis of existing definitions of DT revealed six distinct approaches to the conceptual meaning of this term. The authors conceptually align with the definitions outlined in Group 1, which posits that DT is the confidence in people, processes, and technology to create a secure digital environment, and Group 2, which emphasises trust in data protection, privacy, security, and ethical information handling, thereby supporting a holistic and multidimensional view of this concept. While digital trust entails confidence in the enablers that foster a secure digital environment, data are the most valuable asset in the modern digital world, and its protection, along with users’ privacy, must be included or acknowledged in the definition of the concept. Simultaneously, the authors contest the position concerning the definitions encompassed in Group 3, which asserts that it involves trust in the technical capability and reliability of technology. Although the importance of trustworthy and reliable technology may be a part of the definition, the technology itself, much like any isolated tool, cannot encompass all dimensions of trust-building.

The analysis of the factors that compose the essence of DT confirmed that there are various approaches towards the conceptualisation level of the factors. While some authors define aspects of DT as “ability” and “benevolence” [53], others may go to a very granular level, mentioning “certification” [65] or “prompt resolution of the issues” [61]. Additionally, this study allowed for the identification of an additional category of trust factors—“openness”, that complements Mayer et al.’s [9] previously named and highly recognised three categories—“ability”, “benevolence”, and “integrity”. Such a result confirms the position of Alpcan et al. [3] and Dimitrakos [10], who argued that there is a distinction between trust in traditional and digital contexts, strongly influenced by the situation.

Finally, the proposed definition of DT is valid and represents all identified key factors. Moreover, the new holistic definition performs better from the perspective of key-factor representation and significantly differs from similar definitions of cyber and online trust. Additionally, the operationalised DT key factors may be used in the future for DT measurement.

6. Conclusions

This study addressed the absence of a unified definition of DT by conducting a comprehensive two-phase analysis. First, 42 DT definitions were mapped into six major clusters, illustrating the multidimensional nature of current theoretical discourse on the concept. Second, the study consolidated and extended three recognised trustworthiness dimensions of “ability”, “benevolence”, and “integrity” by introducing “openness” as an additional dimension relevant in digital contexts. From these four categories, nine DT factors were synthesised, including technical (e.g., security and control, secure architecture), relational and user-centric (e.g., competence and performance, user experience), governance and ethics, and transparency-focused (e.g., auditability, traceability) aspects.

Building on these results, the study proposed a new, holistic theoretical DT definition, which was validated through a survey. Compared to other cited definitions, the new definition was shown to capture the essential, multi-layered nature of DT more comprehensively. In particular, the validation results indicated strong to moderate alignment between the definition and all nine DT factors, confirming its conceptual fit.

The proposed definition and DT factors provide researchers and practitioners with a baseline for developing and testing DT measurement instruments. Moreover, organisations that seek to measure, manage, and strengthen DT may use a multi-dimensional view of the concept to enhance trust in their products and services. The expanded perspective on openness, including transparency and independent verification, may assist policymakers in developing and improving the guidance that will represent holistic requirements towards DT and promote blockchain technology usage as an element that facilitates trust.

This paper has several limitations. First, the paper aimed to analyse a scientific body of previous peer-reviewed publications to propose a comprehensive scientific definition of digital trust. Although the DT concept is widely applied in the industry, professional publications were not considered within this analysis due to the focus on the academic understanding and conceptualisation of DT. Second, the authors focused on two scientific databases—Scopus and WoS—to ensure the high credibility of the used sources. Although authors assume that the retrieved sources provided a comprehensive review of the DT concept, adding additional sources, such as IEEE Xplore or Google Scholar, might provide a wider overview of this emerging concept. Third, the paper was focused solely on the DT concept, not taking into consideration and not comparing with the concepts that preceded the development of DT.

Future research on DT may overcome the paper’s focus solely on the academic body of knowledge. By analysing professional literature and industry frameworks, future research may compare how academia and industry understand the concept, and discuss findings with industry experts. This analysis could address the question of the identified factors’ operationalisation within the industry and explore how they can be measured in different contexts (such as e-commerce, finance, and healthcare) and what tools can potentially be employed for this purpose.

Further, future research may address questions about each thematic group of the implications of the definition for various stakeholders, for example, organisations (businesses), consumers, and policymakers. Additionally, a comparison of e-trust, cyber trust, online trust, and DT concepts, understanding their differences and similarities might shed light on the question of whether they represent the same meaning and under what circumstances they are typically used.