Enhancing Trust in Inter-Organisational Data Sharing: Levels of

Assurance for Data Trustworthiness

Florian Zimmer

1 a

, Janosch Haber

2 b

and Mayuko Kaneko

3 c

Industrial Manufacturing, Fraunhofer ISST, Dortmund, Germany

Fujitsu Research of Europe, Slough, U.K.

Data & Security Research Laboratory, Fujitsu Limited, Kanagawa, Japan

ﬂ

Keywords:

Data Trustworthiness, Levels of Assurance, Inter-Organisational Data Sharing, Trust, Data Spaces, Design

Science Research.

Abstract:

With data increasingly acknowledged as a valuable asset, much effort has been put into investigating inter-

organisational data sharing to unlock the value of previously unused data. Hence, research has identiﬁed

mutual trust between actors as essential prerequisite for successful data sharing activities. However, existing

research oftentimes focuses on trust from a data provider perspective only. Our work, therefore, highlights the

unbalanced view of trust and addresses trust barriers from a data consumer perspective. Investigating trust on

a data level, i.e. the assessment and assurance of data trustworthiness, we found that existing solutions focused

on data trustworthiness do not meet the domain requirements of inter-organisational data sharing. This paper

addresses this shortcoming by proposing a new artifact called Levels of Assurance for Data Trustworthiness

(Data LoA) based on a design science research approach. Data LoA provides an overarching, standardised

framework to assure data trustworthiness in inter-organisational data sharing. Our research suggests that the

adoption of this artifact would lead to an increase of data consumer trust. Still, being a ﬁrst iteration artifact,

Data LoA requires further design efforts before it can be deployed.

1 INTRODUCTION

The increasing adoption of information-driven tech-

nology across industries and its integration in nearly

every aspect of life highlights the ever-growing im-

portance of data. Data is considered a central driver

in the acceleration of digital transformation and has

been acknowledged as an essential asset for inno-

vation and growth (Otto et al., 2022). As a result,

inter-organisational data sharing has recently gained

much attention in academia and industry, aiming to

unlock the full potential of previously unused and un-

derutilised data (Tocco and Lafaye, 2022).

However, organisations are often hesitant when

engaging in data sharing activities, with a lack of trust

and transparency mentioned as one of the most funda-

mental barriers (Jussen et al., 2023). The main reason

for the lack of trust mentioned is challenges to data

sovereignty, i.e. the concern of data providers to lose

https://orcid.org/0009-0002-8060-7162

https://orcid.org/0000-0001-5494-9770

https://orcid.org/0000-0001-9873-2557

control over their data once shared with other organ-

isations (von Scherenberg et al., 2024). Data spaces

have emerged to overcome these concerns (Otto et al.,

2022). However, the issue of trust here is predom-

inantly considered from the perspective of the data

provider. We found that current approaches often aim

at preventing a loss of sensitive information or im-

proving data security (Huber et al., 2022), but rarely

mention the data consumers’ risks and their need for

trust in data providers and the data made available by

them (Otto et al., 2022; Tocco and Lafaye, 2022).

Contrarily, data consumers are often referred to

as risk owners, facing potential risks from data

providers’ insufﬁcient measures to ensure data in-

tegrity (ISO and IEC, 2022). Because data is becom-

ing crucial for (automated) decision making (Faheem

Zafar et al., 2017), leveraging (un-)intentionally mod-

iﬁed, incomplete, or compromised data exposes data

consumers to potentially severe consequences from ﬁ-

nancial losses to human harm (Lim et al., 2012; Jai-

girdar et al., 2019). Still, data consumers usually have

no other option than to trust data providers, as trust for

data itself cannot be established otherwise (Alhaqbani

Zimmer, F., Haber, J., Kaneko and M.

Enhancing Trust in Inter-Organisational Data Sharing: Levels of Assurance for Data Trustworthiness.

DOI: 10.5220/0013461800003967

In Proceedings of the 14th International Conference on Data Science, Technology and Applications (DATA 2025), pages 339-346

ISBN: 978-989-758-758-0; ISSN: 2184-285X

339

and Fidge, 2009).

In this paper, we argue that trust in inter-

organisational data sharing should not be limited to

the organisational level but must also encompass the

trustworthiness of the data itself. Following a design

science research (DSR) approach, we review existing

literature on data trustworthiness. We systematically

identify overarching design requirements for existing

solutions and identify implications and challenges of

establishing and assuring data trustworthiness in com-

plex data sharing scenarios. Based on the identiﬁed

problem and solution spaces, we aim to address the

shortcomings of existing artifacts by introducing the

concept of Levels of Assurance for Data Trustworthi-

ness, or Data LoA, a novel framework for enhancing

trust and transparency among data providers and con-

sumers. In its ﬁrst iteration, we develop a founda-

tional model that outlines key actors and their inter-

actions and present a proof of concept (PoC) imple-

mentation which demonstrates our concept. Our main

contributions include:

(i) Compiled design knowledge, identifying exist-

ing work and mapping the problem and solution

space

(ii) A novel Data LoA artifact aimed at enhancing

data consumer trust in data sharing scenarios

The remainder is structured as follows: In Sec-

tion 2 we touch on related work. Section 3 outlines

the research methodology. Section 4 presents the de-

rived design knowledge and proposes the concept of

Data LoA. In Section 5, we discuss implications and

limitations, and future work. Section 6 concludes the

paper.

2 RELATED WORK

2.1 Data Trustworthiness

The trustworthiness of data has been extensively stud-

ied across various domains and applications such as

healthcare, defence, trafﬁc control, and manufactur-

ing (Gomez et al., 2009). It is usually described as

the possibility to ascertain the correctness of data pro-

vided by a data source (Haron et al., 2017). Yet,

a high degree of context and domain dependency so

far have prevented the formulation of a generally ac-

cepted notion of data trustworthiness (Bertino, 2015;

Xu and MacAskill, 2023). Circumnavigating a holis-

tic deﬁnition, literature oftentimes mentions speciﬁc

dimensions of data trustworthiness, most commonly

data quality, availability, security, and compatibility

- each of which concerned with several different as-

pects themselves (Xu and MacAskill, 2023).

Previous work has produced a range of different

artifacts to assure, measure, and deﬁne individual as-

pects of data trustworthiness. For instance, (Ormaza-

bal et al., 2024) present a trust assessment canvas to

gauge the trustworthiness of publicly available med-

ical data. (Foidl and Felderer, 2023) present a trust

score model capable of measuring the trustworthiness

of industrial IoT data sources. And (Leteane and Ay-

alew, 2024) propose a trust-enhancing framework for

data traceability in the context of food supply chains.

Some researchers argue that the variety of ap-

proaches and diversity of deﬁnitions prevents the de-

velopment of an overarching, comprehensive solu-

tion to assure data trustworthiness (Bertino, 2015;

Ebrahimi et al., 2022). Therefore, (Haron et al., 2017)

argue that a combination of different techniques will

be required to fully meet the trust-related require-

ments of data consumers.

2.2 Levels of Assurance

Levels of Assurance (LoA) refer to the degree of con-

ﬁdence that can be assigned to some entity, process,

or system acting or operating as claimed (ISO and

IEC, 2013). LoAs are an assurance technique used

to evaluate and grade complex scenarios, simplifying

and improving decision making and risk management

(Nenadic et al., 2007). More formally, a relying party

utilises provided LoA information to determine their

level of conﬁdence in the credibility of a claimant’s

claim. Usually, there is at least one other party in-

volved, namely the assurance provider, which audits

and assures the claimant’s claim (Mart

ınez-Ferrero

and Garc

ıa-S

anchez, 2018; ISO and IEC, 2013). If

there is no external assurance provider, claims are

self-asserted.

LoAs are mainly used in the domain of identity

validation, for example, in the ISO/IEC 29115

stan-

dard, the eIDAS regulation as proposed by the Euro-

pean Commission, or the NIST 800-63-A

guidelines.

LoAs are deﬁned risk-based, specifying dimensions

of risk that must be addressed and mitigated to as-

sure the credibility of a claim. Hence, the higher the

perceived risk for the relying party, the higher the re-

quired level of conﬁdence in the claim’s validity, and

thus the LoA should be (Mart

ınez-Ferrero and Garc

ıa-

anchez, 2018; ISO and IEC, 2013). A comprehen-

sive LoA concept should guide the claimant on how to

mitigate risks, while providing the relying party with

assurances needed for informed decision making.

https://iso.org/standard/45138.html

https://pages.nist.gov/800-63-3/sp800-63a.html

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

340

Apart from that, eIDAS also aims to improve trust

among adopters by assuring identiﬁcation techniques

and clearly deﬁning liabilities to specify each party’s

responsibilities. Thus, eIDAS certiﬁes commonly

used identiﬁcation techniques with different LoAs,

providing standardised assurances to assess mutual

trust and enabling interoperability in the heteroge-

neous identiﬁcation techniques landscape of the EU

(European Parliament, 2014).

3 METHODOLOGY

This paper aims to address the under-representation of

trust-establishing means for data consumers in inter-

organisational data sharing. To do so, we applied

a rigorous DSR approach following (Peffers et al.,

2007) to design a new Levels of Assurance for Data

Trustworthiness (Data LoA) artifact that provides a

framework for unifying and standardising the assur-

ance of data trustworthiness. The goal of this arti-

fact is twofold: First, it enables data consumers to as-

sess the risks associated with utilising a shared data

asset. Second, it equips data providers with stan-

dardised principles on how to establish different de-

grees of data trustworthiness assurances for the data

they want to share. Together, these mechanisms are

aimed at enhancing trust in inter-organisational data

sharing and enabling interoperability among existing

trust-assuring solutions.

We followed an objective-centred DSR approach

building on existing data trustworthiness assuring ar-

tifacts. However, as previous artifacts were not neces-

sarily developed under DSR, available design knowl-

edge was limited. We therefore started by identifying

the relevant problem and solution spaces, and identi-

ﬁed the challenges, motivations, and goals addressed

by existing artifacts. We did so by conducting a struc-

tured literature review (SLR) as described by (vom

Brocke et al., 2015), deriving design knowledge by

empirical means.

Exploratory Pre-Study

Keyword-based Search n = 318

Title Screenlining n = 177

Abstract Screenlining n = 47

Back- and Forward Search n = 62

Figure 1: Conducted literature search approach.

Our SLR process, illustrated in Figure 1, resulted

in 62 articles. The complete SLR process can be

accessed for full transparency and reproducability at

(Zimmer et al., 2025). We then empirically derived

design knowledge for i) challenges and motivations

in measuring and assuring data trustworthiness, and

ii) common objectives and goals of existing artifacts.

Formalising this information grounded the relevance

of our artifact and focused our design efforts.

We continued our DSR objective-centred ap-

proach as pictured in Figure 2. We derived and se-

lected a set of design objectives aligned with our over-

all goal of enhancing trust in inter-organisational data

sharing, and used these objectives to guide the devel-

opment of our artifact. During this stage, we noticed

that the concept of LoAs for identity validation serves

a very similar purpose in an adjacent domain and de-

cided to use LoAs as inspiration for our artifact, ad-

justing our design goals accordingly.

Step 1

Define

Objectives

Step 2

Design &

Develop-

ment

Step 3

Demon-

stration

Step 4

Evaluation

Step 5

Communi-

cation

Figure 2: The objective-centred DSR approach following

(Peffers et al., 2007) applied in this study.

Based on the identiﬁed design requirements, we

then developed a ﬁrst iteration of the Data LoA arti-

fact. We mainly focused on deﬁning central mecha-

nisms, actors, and their relations to enable early eval-

uation and establish a sound foundation for future it-

erations. We evaluated our artifact by instantiating a

PoC in the context of data spaces to investigate trust

effects in our target domain. Conducting an experi-

mental simulation allowed us to assess the technical

feasibility of our concept and determine limitations

and considerations for future work.

4 RESULTS

4.1 Design Knowledge & Grounding

An objective-centred DSR approach assumes a well-

deﬁned problem space (Peffers et al., 2007). As this

was not the case, we conducted a SLR as described in

Section 3 to extract and derive existing design knowl-

edge. We compiled information on past artifact mo-

tivation and problem space deﬁnition from previous

literature, identifying three main motivations and one

key challenge for providing data trustworthiness as-

surance: i) mitigating undeterminable risks in data

sharing, ii) unlocking the operational value of shared

data assets, and iii) catering to an increasing demand

Enhancing Trust in Inter-Organisational Data Sharing: Levels of Assurance for Data Trustworthiness

341

for trustworthy data. We also identiﬁed the complex-

ity of assessing data trustworthiness as its main chal-

lenge. For brevity, we here only reference key liter-

ature. A full list of the analysed articles and derived

clusters can be found in (Zimmer et al., 2025).

Improving trust in third-party data is crucial due to

the undeterminable risks posed by using untrustwor-

thy data. Data inﬂuences decision-making, impact-

ing accuracy, reliability, and overall business success,

highlighting the operational value of data (Haron

et al., 2017; Karthik and Ananthanarayana, 2016).

Thus, using untrustworthy data can negatively im-

pact business success and lead to severe consequences

(Jaigirdar et al., 2019). Therefore, ensuring data

trustworthiness is driven by improved risk manage-

ment and accountability, using only trustworthy data

in high-risk environments, avoiding low-trustworthy

data (Ardagna et al., 2021).

Furthermore, research suggests that there is an in-

creasing demand for trustworthy data, mentioning an

increased reliance on data for daily operations and the

data-related risk attached to it as the primary cause

(Bertino, 2015; Islam et al., 2025). Additionally, re-

cent work emphasises the increasing demand for re-

liable data in artiﬁcial intelligence (AI) (Anjomshoaa

et al., 2022). However, existing work also highlights

that ensuring and assessing data trustworthiness is a

challenging task. For instance, (Bertino, 2015) argues

that addressing the different facets of data trustwor-

thiness requires a complex combination of different

approaches and techniques.

4.2 Design Objectives

To deﬁne our design objectives, we investigated the

solution space populated by previous research. Based

on our SLR, we identiﬁed 51 previous artifacts related

to improving or assessing data trustworthiness, the

majority of which are in the domain of IoT. Although

the understanding of data trustworthiness differs in

research, many artifacts do appear to have common

goals, which we used to guide our design efforts. In

total, we identiﬁed four central design objectives: i)

improve data consumer trust in shared data assets, ii)

reduce the risk of utilising third-party data assets, iii)

decrease the complexity of assessing data trustwor-

thiness, and iv) enable interoperability of existing ap-

proaches for assessing and assuring data trustworthi-

ness.

Most of the identiﬁed goals are closely tied to

the identiﬁed motivations and are interconnected with

each other. The most important goal of existing solu-

tions is to enhance trust in data, i.e., increasing a data

consumer’s conﬁdence in the data they use (Alkhe-

laiwi and Grigoras, 2015). This is mirrored in re-

search on inter-organisational data sharing, acknowl-

edging trust as the most important factor for it to suc-

ceed (Tocco and Lafaye, 2022). Typically, artifacts

enhance trust by increasing transparency and provid-

ing detailed information, such as data origin or prove-

nance, or by providing a more simpliﬁed aggregated

trust score (Foidl and Felderer, 2023; Leteane et al.,

2024). Therefore, increasing transparency and reduc-

ing trustworthiness assessment complexity are also

goals often mentioned. Moreover, tackling these is-

sues also enables consumers to mitigate risk as they

are enabled to make informed decisions, which ulti-

mately allows them to avoid unsuitable data for use

in potentially high-risk environments. This is in line

with the domain of inter-organisational data sharing,

as many different users with different backgrounds

need to grow conﬁdent in the usage of third-party data

while avoiding costly risks.

A key design objective missing in existing arti-

facts is interoperability. Most solutions address spe-

ciﬁc cases or domains like IoT. Still, existing solu-

tions could play an important role in assuring data

trustworthiness in the context of inter-organisational

data sharing. Yet, there is no standardised trust model

or overarching solution to enhance trustworthiness at

the data level (Ebrahimi et al., 2022). As the iden-

tity LoA eIDAS was introduced for a similar reason,

we believe that the current heterogeneous landscape

of existing data trustworthiness artifacts would bene-

ﬁt from taking an analogous approach (European Par-

liament, 2014). Hence, the idea of LoA could be used

to assess and assure the data trustworthiness based on

existing solutions and measurements in place. There-

fore, we adopt and apply this design goal inspired by

the domain of LoAs for identity veriﬁcation.

4.3 Artifact Description

Based on the formalised design knowledge and de-

rived design goals, we developed a novel frame-

work for data trustworthiness assurance in inter-

organisational data sharing. The resulting artifact is

an LoA-based assurance technique called Levels of

Assurance for Data Trustworthiness, or short Data

LoA. Similar to LoAs in other domains, we deﬁne

Data LoA as follows: Levels of Assurance for Data

Trustworthiness refer to the degree of conﬁdence that

a data asset’s underlying information can be trusted

to be true. In other words, Data LoA ensures the con-

ﬁdence a data consumer can put into a data asset’s

trustworthiness, considering the residual risks related

to aspects not covered by the provided assurance.

Within the Data LoA framework, we propose the fol-

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

342

lowing three actors and interactions as displayed in

Figure 3.

provides claim

Assurance

Provider

audits claimassures claim

Data

Consumer

Data

Provider

Figure 3: Abstract actor model of Data LoA.

Data Consumer. Using the terminology of LoAs

for identity veriﬁcation, the Data Consumer is the re-

lying party. The Data Consumer is the actor who ulti-

mately utilises a given data asset and thus carries the

risk of leveraging non-trustworthy data. In the LoA

framework, the Data Consumer assesses the provided

data trustworthiness assurances to determine whether

to use a speciﬁc data asset, considering the risks asso-

ciated with its use.

Data Provider. The Data Provider is the claimant.

The Data Provider claims that a given data asset pro-

vided by them offers a certain degree of trustworthi-

ness. Speciﬁcally, by providing a certain Data LoA,

the Data Provider claims that appropriate measures

were taken to establish a speciﬁc degree of conﬁdence

in the data asset’s trustworthiness.

Assurance Provider. The Assurance Provider

should be an independent third-party actor who fulﬁls

the role of a trustworthy auditor and assurer. In the

domain of inter-organisational data sharing, it is often

referred to as trust anchor (CEN/WS TDT, 2024). Al-

though not strictly necessary, having independent au-

dits and assurances greatly enhances the level of trust

that a Data Consumer can place in the assured claims

– as self-asserted claims are usually not considered

trustworthy.

Interactions. To establish a Data LoA, the Data

Provider must generate a claim of trustworthiness for

a speciﬁc data asset. This claim is presented to the

Assurance Provider, who conducts an audit of the

given claim. To do so, the Data Provider must provide

sufﬁcient evidence to prove that their trustworthiness

claim indeed holds true. It is then at the discretion

of the Assurance Provider to certify the data asset’s

trustworthiness assurances at a speciﬁc level.

Based on the assured claim and its intended appli-

cation, the Data Consumer can then decide whether to

put their conﬁdence in the trustworthiness of the data

and utilise it - or not. The ﬁnal risk and decision re-

sponsibility still lies with the Data Consumer, but they

now have stronger evidence to inform their decisions

and may have legal grounds to sue Data Providers for

false assurances.

4.4 Demonstration & Evaluation

We evaluated our initial artifact through demonstra-

tion. We followed (Hevner et al., 2004) by opting to

conduct an experimental simulation as well as an in-

formed argument as part of a descriptive evaluation.

According to (Gregor and Hevner, 2013), the evalua-

tion through a PoC is sufﬁcient for novel artifacts. We

implemented our artifact within data spaces to test our

framework for inter-organisational data sharing. This

context allowed us to leverage existing technologies

for scalable, standardised data sharing and examine

the potential impact in practical data sharing scenar-

ios. Our PoC simulates a minimal data space, com-

prising the following three components, as pictured in

Figure 4: data source, a data sink, and a data space.

Data

Consumer

Dataspace

«EDC»

Connector

Negotiation

& Transfer

«EDC»

Connector

Data

Provider

«Backend»

Data Sink

«Backend»

Data Source

LoA

decide usage

Data Data

LoA

select, publish

Figure 4: Data LoA PoC experimental setup.

The data space is comprised of two data space

connectors, which enable a sovereign data exchange

between a data provider and consumer. They of-

fer features for data discovery, policy negotiation,

and data transfer. The connectors were implemented

using the community-driven open-source framework

Eclipse Data Space Components

. The Data Source

and Data Sink are simple Python backends aiming to

provide or accept dummy data over a REST API. In

this PoC, we focus solely on the interaction between

the data provider and data consumer to reduce com-

plexity. As a result, all claims are considered self-

asserted.

The PoC is deployed using Docker on a virtual

machine running Linux Ubuntu. With everything in

place, the following steps are performed. First, the

provider selects a dataset from their data source to be

https://projects.eclipse.org/projects/technology.edc

Enhancing Trust in Inter-Organisational Data Sharing: Levels of Assurance for Data Trustworthiness

343

published in the data space. The published asset is

part of a data catalog and describes both the dataset

and the usage policies associated with it. In addition,

the data provider can include any other information in

the data catalog, which, in our case, includes the Data

LoA claim generated and associated with the data as-

set.

Then, the consumer uses their connector to re-

quest the provider’s catalog and to inspect the regis-

tered assets. Typically, a consumer decides whether

to request and use a given asset based on the infor-

mation provided in the catalog. However, this infor-

mation is oftentimes rather sparse and only contains a

general description of the data. When using the Data

LoA framework, the data consumer is provided with

an additional, comprehensive and standardised Data

LoA claim. In this demonstration, we assume the

provided claim to be acceptable. Therefore, the con-

sumer decides to request and negotiate the data offer

upon which the data is ﬁnally transferred.

5 DISCUSSION

In this paper, we present a novel framework for as-

suring the trustworthiness of data to address the trust

deﬁcit of data consumers in inter-organisational data

sharing. To establish a comprehensive framework,

we followed an objective-oriented DSR approach and

identiﬁed LoAs as a promising foundation. We then

developed a ﬁrst iteration of our novel Data LoA con-

cept and demonstrated its feasibility through a PoC

implementation. We suggest that our work provides

the following contributions:

Formalisation of Design Knowledge. To the best

of our knowledge, this is the ﬁrst attempt to explic-

itly state existing design knowledge and objectives for

data trustworthiness artifacts. Using a SLR, we iden-

tiﬁed three main motivations and one main challenge.

We suggest that our work provides a sound foundation

for future DSR-based contributions, and we hope that

it will lead to the development of new and improved

artifacts for data trustworthiness. Our Data LoA is

the ﬁrst contribution to beneﬁt from this formalisa-

tion, presenting a novel artifact to address identiﬁed

challenges in a more comprehensive manner.

Proposal of Data LoA artifact. Based on the

identiﬁed problem and solution space, we designed

an overarching data trustworthiness assurance frame-

work inspired by existing LoAs in the identity do-

main. Our experimental simulation demonstrated

how the Data LoA claim can be presented and ex-

changed in a typical inter-organisational data sharing

environment. We suggest that our solution increases

transparency, thereby promoting trust for consumers

and enabling them to make sound decisions. This ul-

timately decreases risk, as consumers are enabled to

utilise only data which matches their demands, based

on a simple claim.

5.1 Limitations & Future Work

Despite conducting a rigorous design approach, our

study is subject to limitations. First, the design knowl-

edge was derived using an SLR approach. We at-

tempted to uncover missing relevant literature by con-

ducting backward and forward searches as part of the

SLR. Still, there remains the possibility of unidenti-

ﬁed relevant related work.

Second, our artifact is in an early stage. We

haven’t speciﬁed assurance levels yet, as a holistic

deﬁnition of data trustworthiness must be developed

ﬁrst. We chose to publish this iteration to encourage

further work and establish a foundation for future de-

velopments. Still, our concept was carefully validated

using an experimental simulation.

Third, although interoperability is a design goal,

the initial Data LoA concept does not yet address it.

It focuses instead on reducing risk, improving trust,

and lowering assessment complexity, as noted in prior

work. Nonetheless, by adopting the idea of LoA to

data trustworthiness, we believe future iterations will

tackle this issue.

Given these limitations, we suggest the following

future work: First, without a common understanding

of data trustworthiness, deﬁning speciﬁc LoA levels

and ensuring user comprehension remains challeng-

ing. Recent standardisation efforts by the CEN work-

ing group Trusted Data Transaction could provide a

promising starting point. (CEN/WS TDT, 2024).

Second, we recommend conducting more DSR

cycles to advance the Data LoA framework. In partic-

ular, deﬁning speciﬁc levels is essential for providers

to make accurate claims and for consumers to assess

risk properly. We suggest deriving these deﬁnitions

from ongoing work on risk dimensions and attack tax-

onomies, as existing LoAs in other domains typically

focus on risk. Other aspects of data trustworthiness,

such as the reputation of the source or country of ori-

gin, should also be considered.

Finally, we suggest identifying relevant domains,

drivers for adoption, and potential implementation

challenges. This ensures our framework gains wide

adoption by clearly communicating target applica-

tions, beneﬁts, and trade-offs. For example, (He et al.,

2015; Hou et al., 2024) note trade-offs between im-

proving data trustworthiness and factors like cost or

privacy. Based on our current understanding, relevant

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

344

domains include critical infrastructure or automated

systems in sensitive domains such as healthcare or de-

fense. However, AI could also greatly beneﬁt from

leveraging data trustworthiness assessments, e.g., to

assign different weights to training data based on their

Data LoA levels, improving accuracy and reliability.

6 CONCLUSION

This paper presents the novel concept of LoA for

data trustworthiness. Data LoA provides a standard-

ised framework to ensure data trustworthiness, ad-

dressing the trust deﬁcit of data consumers in inter-

organisational data sharing. It aims to enhance trust,

reduce risks of using shared data, simplify the assess-

ment of data trustworthiness, and enable interoper-

ability between existing and new approaches. These

goals arise from a rigorous DSR approach and the for-

malisation of existing design knowledge.

We found that although our Data LoA concept

addresses most of the identiﬁed objectives in theory,

more work is needed to deploy it in real scenarios. Es-

pecially the goal of interoperability, the level deﬁni-

tion and its implementation need more attention. We

suggest that further DSR cycles should be performed

to incrementally enhance the Data LoA artifact pre-

sented here.

Our work contributes to the ﬁeld of design re-

search for data trustworthiness by formalising design

knowledge and presenting a new artifact. We encour-

age researchers to utilise our ﬁndings as a founda-

tion to further investigate the subject. We hope that

our contributions help to increase research efforts ad-

dressing the identiﬁed shortcomings in consumer trust

and ultimately improve and extend data sharing activ-

ities across organisations.

ACKNOWLEDGEMENTS

CRediT author statement: Florian Zimmer: Con-

ceptualisation, Methodology, Software, Validation,

Investigation, Writing - Original Draft, Visualisation.

Janosch Haber: Conceptualisation, Writing - Re-

view & Editing, Mayuko Kaneko: Conceptualisa-

tion, Writing - Review & Editing, Project adminis-

tration.

REFERENCES

Alhaqbani, B. and Fidge, C. (2009). A time-variant med-

ical data trustworthiness assessment model. In 2009

11th International Conference on e-Health Network-

ing, Applications and Services (Healthcom), pages

130–137.

Alkhelaiwi, A. and Grigoras, D. (2015). The origin and

trustworthiness of data in smart city applications. In

Proceedings of the 8th International Conference on

Utility and Cloud Computing, UCC ’15, pages 376–

382. IEEE Press.

Anjomshoaa, A., Elvira, S. C., Wolff, C., P

erez Ba

un, J. C.,

Karvounis, M., Mellia, M., Athanasiou, S., Katsifodi-

mos, A., Garatzogianni, A., Tr

ugler, A., Serrano, M.,

Zappa, A., Glikman, Y., Tuikka, T., and Curry, E.

(2022). Data platforms for data spaces. In Curry,

E., Scerri, S., and Tuikka, T., editors, Data Spaces,

Springer eBook Collection, pages 43–64. Springer In-

ternational Publishing and Imprint Springer, Cham.

Ardagna, C. A., Asal, R., Damiani, E., Ioini, N. E., Elahi,

M., and Pahl, C. (2021). From trustworthy data to

trustworthy iot: A data collection methodology based

on blockchain. ACM Trans. Cyber-Phys. Syst., 5(1).

Bertino, E. (2015). Data trustworthiness—approaches and

research challenges. In Garcia-Alfaro, J., Herrera-

Joancomart

ı, J., Lupu, E., Posegga, J., Aldini, A.,

Martinelli, F., and Suri, N., editors, Data Privacy

Management, Autonomous Spontaneous Security, and

Security Assurance, volume 8872 of Lecture Notes

in Computer Science, pages 17–25. Springer Interna-

tional Publishing, Cham.

CEN/WS TDT (July 2024). Trusted data transaction: Part

1: Cwa 18125.

Ebrahimi, M., Tadayon, M. H., Haghighi, M. S., and Jol-

faei, A. (2022). A quantitative comparative study of

data-oriented trust management schemes in internet of

things. ACM Trans. Manage. Inf. Syst., 13(3).

European Parliament (23 July / 2014). Regulation no

910/2014 on electronic identiﬁcation and trust ser-

vices fro electronic transactions in the internal market

and repealing directive: eidas regulation.

Faheem Zafar, Abid Khan, Saba Suhail, Idrees Ahmed,

Khizar Hameed, Hayat Mohammad Khan, Farhana

Jabeen, and Adeel Anjum (2017). Trustworthy data:

A survey, taxonomy and future trends of secure prove-

nance schemes. Journal of Network and Computer

Applications, 94:50–68.

Foidl, H. and Felderer, M. (2023). An approach for assess-

ing industrial iot data sources to determine their data

trustworthiness. Internet of Things, 22:100735.

Gomez, L., Laube, A., and Sorniotti, A. (2009). Trustwor-

thiness assessment of wireless sensor data for business

applications. In 2009 International Conference on

Advanced Information Networking and Applications,

pages 355–362.

Gregor, S. and Hevner, A. R. (2013). Positioning and pre-

senting design science research for maximum impact.

MIS Quarterly, 37(2):337–355.

Haron, N., Jaafar, J., Aziz, I. A., Hassan, M. H., and

Shapiai, M. I. (2017). Data trustworthiness in in-

ternet of things: A taxonomy and future directions.

In 2017 IEEE Conference on Big Data and Analytics

(ICBDA), pages 25–30.

Enhancing Trust in Inter-Organisational Data Sharing: Levels of Assurance for Data Trustworthiness

345

He, D., Chan, S., and Guizani, M. (2015). User privacy and

data trustworthiness in mobile crowd sensing. IEEE

Wireless Communications, 22(1):28–34.

Hevner, A. R., March, S. T., Park, J., and Ram, S. (2004).

Design science in information systems research. MIS

Quarterly.

Hou, C., Zhou, C., Wu, C. G., Cong, R., and Li, K. (2024).

Optimization of cloud-based multi-agent system for

trade-off between trustworthiness of data and cost of

data usage. IEEE Transactions on Automation Science

and Engineering, 21(1):106–122.

Huber, M., Wessel, S., Brost, G., and Menz, N. (2022).

Building trust in data spaces. In Otto, B., ten Hompel,

M., and Wrobel, S., editors, Designing Data Spaces,

Springer eBook Collection, pages 147–164. Springer

International Publishing and Imprint Springer, Cham.

Islam, M. M., Karmakar, G. C., Kamruzzaman, J., Mur-

shed, M., and Chowdhury, A. (2025). Trustworthiness

of iot images leveraging with other modal sensor’s

data. IEEE Internet of Things Journal, 12(1):163–173.

ISO and IEC (April, 2013). Information technology — se-

curity techniques — entity authentication assurance

framework: Iso/iec 29115.

ISO and IEC (August 2022). Information security, cyberse-

curity and privacy protection — evaluation criteria for

it security - part 5: Iso/iec 15408-5:2022.

Jaigirdar, F. T., Rudolph, C., and Bain, C. (2019). Can i

trust the data i see? a physician’s concern on medi-

cal data in iot health architectures. In Proceedings of

the Australasian Computer Science Week Multiconfer-

ence, ACSW ’19, New York, NY, USA. Association

for Computing Machinery.

Jussen, I., Schweihoff, J., and M

oller, F. (2023). Tensions

in inter-organizational data sharing: Findings from lit-

erature and practice. In 2023 IEEE 25th Conference

on Business Informatics (CBI), pages 1–10. IEEE.

Karthik, N. and Ananthanarayana, V. S. (2016). Sensor

data modeling for data trustworthiness. In 2016 IEEE

Trustcom/BigDataSE/ISPA, pages 909–916.

Leteane, O. and Ayalew, Y. (2024). Improving the trust-

worthiness of traceability data in food supply chain

using blockchain and trust model. The Journal of The

British Blockchain Association, 7(1):1–12.

Leteane, O., Ayalew, Y., and Motshegwa, T. (2024). A

multi-package trust model for improving the trustwor-

thiness of traceability data in blockchain-based beef

supply chain. In IEEE Conference on Dependable and

Secure Computing, pages 155–162.

Lim, H. S., Ghinita, G., Bertino, E., and Kantarcioglu, M.

(2012). A game-theoretic approach for high-assurance

of data trustworthiness in sensor networks. In 2012

IEEE 28th International Conference on Data Engi-

neering, pages 1192–1203.

Mart

ınez-Ferrero, J. and Garc

ıa-S

anchez, I.-M. (2018). The

level of sustainability assurance: The effects of brand

reputation and industry specialisation of assurance

providers. Journal of Business Ethics, 150(4):971–

990.

Nenadic, A., Zhang, N., Yao, L., and Morrow, T. (2007).

Levels of authentication assurance: an investigation.

In Third International Symposium on Information As-

surance and Security, pages 155–160. IEEE.

Ormazabal, A., Berry, D., and Hederman, L. (2024).

Co-development of a tool to help clinicians decide

upon the trustworthiness of patient generated health

data. In 2024 IEEE 37th International Symposium

on Computer-Based Medical Systems (CBMS), pages

442–449.

Otto, B., ten Hompel, M., and Wrobel, S., editors (2022).

Designing Data Spaces: The Ecosystem Approach

to Competitive Advantage. Springer eBook Collec-

tion. Springer International Publishing and Imprint

Springer, Cham, 1st ed. 2022 edition.

Peffers, K., Tuunanen, T., Rothenberger, M. A., and Chat-

terjee, S. (2007). A design science research method-

ology for information systems research. Journal of

Management Information Systems, 24(3):45–77.

Tocco, F. and Lafaye, L. (2022). Data platform solutions.

In Otto, B., ten Hompel, M., and Wrobel, S., editors,

Designing Data Spaces, Springer eBook Collection,

pages 383–393. Springer International Publishing and

Imprint Springer, Cham.

vom Brocke, J., Simons, A., Riemer, K., Niehaves, B., Plat-

tfaut, R., and Cleven, A. (2015). Standing on the

shoulders of giants: Challenges and recommendations

of literature search in information systems research.

Communications of the Association for Information

Systems, 37.

von Scherenberg, F., Hellmeier, M., and Otto, B. (2024).

Data sovereignty in information systems. Electronic

Markets, 34(1).

Xu, J. and MacAskill, K. (2023). A carbon data trustwor-

thiness framework for the construction sector. In Pro-

ceedings of the 2023 European Conference on Com-

puting in Construction and the 40th International CIB

W78 Conference. European Council for Computing in

Construction.

Zimmer, F., Haber, J., and Kaneko, M. (2025). Enhanc-

ing trust in inter-organisational data sharing: Levels

of assurance for data trustworthiness - literature body.

Zenodo, https://doi.org/10.5281/ zenodo.14639350.

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

346