Navigating Responsible AI Adoption

Daniela Oliveira

Independent Researcher, Canada

Keywords: Responsible Artificial Intelligence, Responsible Adoption, Artificial Intelligence Governance, Change

Management, Knowledge Management, Organizational Learning.

Abstract: Responsible Artificial Intelligence has been a largely discussed topic among organizations that develop or are

aiming to regulate Artificial Intelligence (AI) solutions. Much less attention has been given to organizations

willing to adopt AI in a responsible manner. Organizations that do not develop AI need practical guidance on

how to implement Responsible AI principles. This contribution outlines the challenges organizations face

integrating the Responsible AI paradigm and suggests some solutions.

1 INTRODUCTION

One of the most distinguishing characteristics of

Artificial Intelligence (AI) technologies from other

kinds of technological solutions is how intrinsically

connected to data these technologies are. This

characteristic brings a different dimension to software

development: customizing an AI solution to a specific

context is highly dependent on this context (Friedler

et al., 2021).

As a consequence, the impact and risks of an out-

of-the-lab solution can be extremely different from

the impact and risks of the implemented solution. The

fact that AI technologies mostly automate cognitive

processes solidifies this condition. Adopting an AI

solution is not only about training workforce to use a

new technological solution, but also reaching a point

where the workforce thinking is enhanced or

replicated in a satisfactory manner - and not

diminished - by the technological solution. A point

found in the middle ground between adapting the AI

solution to the context of application and integrating

humans to the development process. The

development and adoption processes of the

technological solution are then rather close,

sometimes even overlapping, when it comes to AI,

compared to traditional technologies.

AI solutions have additional challenges: they

need a conscious effort to predict and reduce the harm

they can cause and become what is known as a

Responsible AI solution (Celdran et al., 2023; Siala

https://orcid.org/0000-0001-9285-0173

& Wang, 2022; Université de Montréal, 2018). The

development or adoption processes have also to

consider the possible harms the AI solutions can

provoke.

Considering the harm an AI solution can provoke

has its share of context-specific considerations, as

each organization has its own culture and processes,

influenced by national and regional culture, market

and regulatory practices, to mention only some of the

aspects that compose the success or failure of

organizations. AI adoption can impact much more

than the technology infrastructure or the data

management practices of an organization. It can

motivate changes in human resources practices

(Tursunbayeva & Renkema, 2022), brand, reputation

(World Economic Forum, n.d.) and knowledge

management (Jarrahi et al., 2022), to name a few, and

present issues throughout the lifecycle of a product or

service. AI solutions can motivate situations where

unexpected human behavior in processes where AI

solutions are integrated produce unexpected

outcomes, possibly leading to physical harm; they can

exacerbate gaps in workplace training; motivate

unclear identification of human and AI’s work, with

mismatching expectations and evaluation practices;

monitoring mechanisms that are centered in AI

solutions and disregard their interaction with

employees and other humans; policies that were not

updated to reflect the complexity of collaborating

with AI solutions; unaligned human / AI quality

assurance initiatives; and non-optimal timeliness.

Oliveira, D.

Navigating Responsible AI Adoption.

DOI: 10.5220/0012262600003598

In Proceedings of the 15th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2023) - Volume 2: KEOD, pages 339-345

ISBN: 978-989-758-671-2; ISSN: 2184-3228

339

AI solutions have not only to be customized in

terms of what functionalities they should address, but

also how these functionalities should be addressed by

the solution. The remaining portrait is that no two AI

implemented solutions are alike in terms of impacts

and risks and should be even less in their adoption

processes.

Adoption processes are led by the organization

willing to successfully integrate a technological

solution in their operations or strategic activities. In

AI solutions, they can represent more than half of the

analytics budget (Fountaine et al., 2019). However,

little attention has been granted to AI adoption

processes focusing on Responsible AI principles,

making it difficult for organizations to plan resources

in order to responsibly adopt an AI solution.

This paper aims to shed light on adoption

processes, and the number and order of challenges

involved in these processes. It is organized as follows:

this Introduction, an overview of the Responsible AI

paradigm in section 2 and of challenges concerning

the adoption of AI solutions in section 3; a definition

of the Responsible AI Adoption process in section 4;

an overview of the Responsible AI Adoption process

challenges in section 5, followed by the conclusion of

the contribution.

2 RESPONSIBLE AI

Responsible AI is a term often associated with a

global movement to ensure AI solutions’ risks are

addressed and mitigated. Responsible AI technology

has been defined as fair and accountable (Agarwal &

Mishra, 2021; Siala & Wang, 2022), explainable

(Agarwal & Mishra, 2021), respecting privacy and

fostering transparency (Siala & Wang, 2022; B. K.

Vassileva, 2021), trustworthiness and empathy (Siala

& Wang, 2022). Responsible AI guidelines such as

the Montreal Declaration for a Responsible

Development of AI (Université de Montréal, 2018)

and the Responsible Microsoft Standard (Microsoft

Corporation, 2022) have been created by government

bodies, AI development companies and research

institutes, to name a few. Lukkien et al. (2023)

recognize a “growing prevalence of frameworks,

principles, and guidelines to inform responsible AI

innovation” (p.156) but deplore that most of them

present high-level principles with excessive room for

interpretation along with limited practical measures

in specific context of use.

3 AI ADOPTION

AI has been the object of unprecedent technological

enthusiasm, to the fact that organizations are willing

to make drastic changes in favour of its adoption.

Alignment between business and information

technology objectives was one of the top concerns

two decades ago (Reich & Benbasat, 2000), but it has

been argued by an AI development company that

“companies must break down organizational and

cultural barriers that stand in AI’s way” (Fountaine et

al., 2019). In practice, how much change should an

organization absorb in favour of AI adoption to

remain competitive? This is a question highly

bounded to the organizational context.

The AI adoption process must be adapted to

reflect an organization’s regional position, business

domain or the challenges of a specific kind of AI

technology. The AI adoption process might range

from stablishing key performance indicators that are

meaningful to the organization to fostering cultural

changes, passing by scrutinizing technical and

business interoperability potentially affected by the

AI solution adoption.

The AI adoption process must also be adapted to

reflect the needs surrounding organizational culture

regarding technology. Humans may trust

technological solutions more than they should, a

phenomenon known as cognitive complacency

(Jarrahi, 2019), even when their outputs seem to be

wrong or inadequate, a situation of automation bias

(Skitka et al., 2000). This factor may be more

prominent in some organizations than in others

(Alon-Barkat & Busuioc, 2023). In addition, great

appetite for AI may lead to overly confident attitudes

(Perry et al., 2022) that make for longer turnaround

reactions. Without organizational culture changes,

the collaboration between humans and AI may

generate deceiving results.

In addition, AI solutions implementation must be

preceded by a risk assessment analysis (Brand, 2022;

Cebulla et al., 2022; Clarke, 2019; Leijnen et al.,

2020; Nagbøl et al., 2021; Oliveira & Dalkir, 2022;

Qiang et al., 2023). Risks may be inherent to the

solution, or a consequence of the application of the

solution in the use context. Concerning population

characteristics, an example of inherent risk of a

solution is one that discriminates a portion of the

population (Mattu, 2016), while one that is a

consequence of the application in the use context is

the difference between the population characteristics

in the time range used for training the model and

when the model was put to use (Suresh & Guttag,

2021). Risks inherent to solutions can be analysed by

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

340

developers and academic actors, but risks that stem

from the application of the solution in the use context

depend on an analysis of the use context against the

documentation of the AI solution following

Responsible AI principles (Mitchell et al., 2019),

which may vary from one organization to another.

After the implementation of an AI solution,

monitoring the output of the solution and engagement

levels among the direct and indirect user population

helps ensure the predicted return on investment is

realized without loss of client base or the addition of

excessive restrictions to operations. The exercise of

identifying what to monitor, however, has to begin

before an AI solution is implemented.

4 RESPONSIBLE AI ADOPTION

Much attention has been given to Responsible AI

solutions development which, arguably, must

integrate different kinds of stakeholders, according to

practitioners, academicians and vendors alike

(Minkkinen et al., 2023; Obermeyer et al., 2021).

However, organizations that do not develop AI

systems seem to have been excluded from the

discussion. An analysis of Responsible AI studies and

frameworks yielded only three types of stakeholders:

individuals and national or international bodies and

organizations involved with AI regulation:

technology companies, professional bodies or

research institutes (Deshpande & Sharp, 2022). This

situation creates a gap between Responsible AI in

theory and in practice. While AI can be the focus of

activity of academicians and vendors, it is not the case

for most organizations willing to adopt AI. These

organizations may have less knowledge of AI than

developers might assume (Richards et al., 2020) but

are, nevertheless, an important element in the AI

environment. They are the organizations holding the

data used to train AI models for specific tasks, the

organizations offering AI-based services to

individuals and other organizations, the organizations

that allocate financial and human resources for the

acquisition of AI systems and the first organizations

to be subject to negative operational, reputational,

relational, and legal impacts of AI if they were to take

place. These organizations need clear substantiation

as to how Responsible AI principles can be flexibly

attuned in context (Bærøe et al., 2020; Lukkien et al.,

2023). Guiding principles can evolve an

organization’s thinking on Responsible AI, “but they

are not sufficient for implementing responsible AI

principles across everything from development to

acquisition to operations” (Probasco, 2022, p. 1).

Due to the multifaceted impact of AI, much of the

Responsible AI gains can only be achieved if

followed by a Responsible AI Adoption process. In

order to operationalize the adoption of AI solutions

respecting responsible principles, Leijnen et al.

(2020) invoke the importance of assessing AI

solutions before implementation, and the inclusion of

usability principles and agile approaches. Adopting

AI in a responsible manner means that practical

aspects of the adoption were analysed and accounted

for in the decision to adopt the AI solution. It may

mean as little organizational reflection as ensuring the

solution is used as per recommended by its

developers (Mitchell et al., 2019) or as much as

triggering business processes; data visualisation

initiatives; stakeholders participation; strategic

positioning and information technology architecture

analysis and evaluation, among other possibilities.

Some of the simple aspects of Responsible AI

Adoption are related to communication and feedback:

removing obstacles for people to voice concerns over

a specific solution. However, organizations need to

reflect on who, apart the user population, should have

the ability to voice concerns over a specific solution.

Employees and managers may contribute in decisive

ways not otherwise considered (Rolls Royce, 2021)

and change management, particularly participative,

and effective integration of domain knowledge have

been correlated with successful AI adoption (von

Richthofen et al., 2022). Organizations also need to

reflect on how to conduct this input collection and

treatment.

Some of the more complex Responsible AI

Adoption aspects are related to the organization’s

positioning in the AI environment. For example, how

valuable are the organization's data to AI providers?

With this understanding, some organizations

negotiated reciprocal agreements that consider the

value of the data involved (Siala & Wang, 2022).

How far have potential providers of an AI solution

adopted Responsible AI principles? Responsible AI

Adoption may help envision responsible initiatives to

palliate shortcomings of an AI solution. For example,

where the replication of bias present in the training

data can be an issue (Au Yeung et al., 2023), the

Responsible AI Adoption process can include

retraining the solution with data that does not contain

harmful biases and a proof-of-concept decisive stage.

The Responsible AI Adoption can be a

differential for organizations enforcing a growth of

their AI solutions portfolio aligned with Responsible

AI values. It can also place the organization a step

forward in regions where regulations are being

developed to hold organizations adopting AI

Navigating Responsible AI Adoption

341

accountable of the due diligence implied in the

Responsible AI paradigm.

5 RESPONSIBLE AI ADOPTION

CHALLENGES

5.1 Selecting Responsible AI

Frameworks and Principles

A Responsible AI Adoption strategy should reflect

the principles of Responsible AI applicable and

contain actionable measures in the context of use.

While the number of Responsible AI guidelines

containing principles has been growing, the number

of guidelines allowing for operationalization of

Responsible AI are still scarce (Lehoux et al., 2023;

Lukkien et al., 2023; Narayanan & Schoeberl, 2023).

In a study that analyses the crossroads between

long-term care, AI and responsible innovation,

Lukkien et al. (2023) argue that Responsible AI can

only be fostered with practical measures that apply

the principles conveyed. The study aimed to identify

concrete measures to influence design and/or

implementation of actual AI solutions in a specific

context of use. The kind of guidance the authors

sought were named “process-based frameworks” by

Narayanan and Schoeberl (2023): frameworks

offering a blueprint helping organizations to prioritize

aspects of system design, identify accountability

lines, establish the infrastructure, resources and

capabilities needed to operationalize Responsible AI.

These frameworks fall under the category of

"operational tools" (Lehoux et al., 2023), documents

that provide in-depth hands-on guidance on a

particular issue with detailed explanations, real-word

examples, step-by-step activities and further

resources.

A portrait of the challenge faced by organizations

in order to benefit from Responsible AI Adoption is

reflected in the research effort of the study of Lukkien

et al. (2023): from 3,339 documents advocating for

Responsible AI, only 8 contained practical measures

in the context of use. Current awareness of

Responsible AI guidelines can be time-consuming.

Process-based guidelines are more sensitive to the

context of use than Principles-based guidelines. In

2021, there were at least 170 frameworks or tools to

support Responsible AI operationalization

(Deshpande & Sharp, 2022). The number of process-

based guidelines is growing, but the challenge to

select and apply these tools persist (Lehoux et al.,

2023; Narayanan & Schoeberl, 2023). The work of

Narayanan and Schoeberl (2023) eases the

Responsible AI Adoption process in an interesting

way. The authors created a taxonomy to help

organizations navigate process-based Responsible AI

frameworks based on the study of 45 generic

Responsible AI frameworks. The Matrix for Selecting

Responsible AI Frameworks was created to assist

organizations identifying frameworks that meet their

specific needs.

However, in contexts where process-based

frameworks are not available, the need to evaluate

impact and mitigate risks are still present. An ethics

framework, guiding the interpretation and translation

of Responsible AI principles into actionable measures

is therefore helpful. Principles-based frameworks

have great variance in scope (Fjeld et al., 2020; Jobin

et al., 2019; Lehoux et al., 2023), even though some

consensus can be verified (Morley et al., 2020). The

context of use of the AI solution might also be the

object of more than one principles-based guideline.

An ethics framework can guide the organization in

the interpretation and translation of more than one

principles-based guideline.

5.2 Finding the Right Ethics

Framework

Applying ethics to business has always been a

delicate endeavour (Murray, 1997) and it is no

different when it comes to AI. An ethics framework

that is aligned with the organizational culture can help

identify risks, prioritize initiatives, obtain buy-in,

guide training and communication around the

adoption of AI and help the interpretation of

principles-based guidelines and their translation into

actionable measures. Ethics framework do not

address AI challenges themselves. Instead, they offer

ways to approach AI challenges and solutions.

For instance, Verbeek and Tijink (2020) argue for

a three-step approach: 1) considering the technology

in context; 2) involving actors, values and effects; 3)

identifying options for action, which can be: a) co-

creation with users and b) ethics by design, in context

and in use.

The co-creation approach is also preconized by

Bruneault et al. (2022). This framework specifies

individual, organizational and social attitudes and

structure. All the actors involved, the authors argue,

should continuously question pre-conceived ideas,

should employ the nuancing that characterizes

concrete applications and should recognize

knowledge limitations at any given moment.

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

342

5.3 Informing AI Governance

Effective AI governance should shape reality

according to governing concepts and also empower

agents of this reality to shape their governing

concepts (Noiseau, 2023). This two-way

interdependence can be used as a way to avoid

effective accountability (Floridi, 2019), but coupled

with a Responsible AI Adoption process, motivated

by actual applications and aligned with the

organizational culture, can be an effective path to

actually implement and maintain Responsible AI

solutions. The measurable performance indicators

and clear criteria for monitoring risks resulting from

the Responsible AI Adoption process can contribute

for a more effective AI governance.

6 CONCLUSION

To be effective, the Responsible AI paradigm

demands guidelines that are both broad in coverage

and specific in advice, as well as the identification of

impact and risks of the adoption of AI solutions in the

domain and context of use. Whereas articulating the

Responsible AI paradigm is a task better performed

by academia and regulatory bodies, the identification

of actionable measures with little room for

interpretation demands specific domain and

organizational knowledge.

This contribution outlined activities related to the

development and implementation of AI in

organizations in order to foster a responsible,

continuous, incremental and aligned AI adoption.

These activities are impacted by the Responsible AI

paradigm but aim actionable measures, adapted to the

domain and context of use of the AI solution. This

contribution suggests the need for talent, time, budget

and particular expertise to promote Responsible AI

Adoption processes.

REFERENCES

Agarwal, S., & Mishra, S. (2021). Responsible AI:

Implementing ethical and unbiased algorithms.

Springer.

Alon-Barkat, S., & Busuioc, M. (2023). Human-AI

Interactions in Public Sector Decision Making:

‘Automation Bias’ and ‘Selective Adherence’ to

Algorithmic Advice. Journal of Public Administration

Research and Theory, 33(1), 153–169.

https://doi.org/10.1093/jopart/muac007

Au Yeung, J., Kraljevic, Z., Luintel, A., Balston, A., Idowu,

E., Dobson, R. J., & Teo, J. T. (2023). AI chatbots not

yet ready for clinical use. Frontiers in Digital Health,

5. https://doi.org/10.3389/fdgth.2023.1161098

Bærøe, K., Miyata-Sturm, A., & Henden, E. (2020). How

to achieve trustworthy artificial intelligence for health.

Bulletin of the World Health Organization, 98(4), 257–

262. https://doi.org/10.2471/BLT.19.237289

Brand, D. J. (2022). Responsible Artificial Intelligence in

Government: Development of a Legal Framework for

South Africa. eJournal of eDemocracy and Open

Government, 14(1), 130–150. Scopus. https://doi.

org/10.29379/jedem.v14i1.678

Bruneault, F., Laflamme, A. S., & Mondoux, A. (2022).

Former à l’éthique de l’IA en enseignement supérieur:

Référentiel de compétence. SocArXiv. https://doi.org/

10.31235/osf.io/38tfv

Cebulla, A., Szpak, Z., Howell, C., Knight, G., & Hussain,

S. (2022). Applying ethics to AI in the workplace: The

design of a scorecard for Australian workplace health

and safety. AI & SOCIETY. https://doi.org/

10.1007/s00146-022-01460-9

Celdran, A. H., Kreischer, J., Demirci, M., Leupp, J.,

Sanchez, P. M., Franco, M. F., Bovet, G., Perez, G. M.,

& Stiller, B. (2023). A Framework Quantifying

Trustworthiness of Supervised Machine and Deep

Learning Models. 2938–2948.

Clarke, R. (2019). Principles and business processes for

responsible AI. Computer Law and Security Review,

35(4), 410–422. Scopus. https://doi.org/10.

1016/j.clsr.2019.04.007

Deshpande, A., & Sharp, H. (2022). Responsible AI

Systems: Who are the Stakeholders? AIES 2022 -

Proceedings of the 2022 AAAI/ACM Conference on AI,

Ethics, and Society, 227–236. Scopus. https://doi.org/

10.1145/3514094.3534187

Fjeld, J., Achten, N., Hilligoss, H., Nagy, A., & Srikumar,

M. (2020). Principled Artificial Intelligence: Mapping

Consensus in Ethical and Rights-Based Approaches to

Principles for AI (SSRN Scholarly Paper 3518482).

https://doi.org/10.2139/ssrn.3518482

Floridi, L. (2019). Translating Principles into Practices of

Digital Ethics: Five Risks of Being Unethical.

Philosophy & Technology, 32(2), 185–193.

https://doi.org/10.1007/s13347-019-00354-x

Fountaine, T., McCarthy, B., & Saleh, T. (2019). Building

the AI-Powered Organization: Technology isn’t the

biggest challenge. Culture is. Harvard Business

Review.

Friedler, S. A., Scheidegger, C., & Venkatasubramanian, S.

(2021). The (Im)possibility of fairness: Different value

systems require different mechanisms for fair decision

making. Communications of the ACM, 64(4), 136–143.

https://doi.org/10.1145/3433949

Jarrahi, M. H. (2019). In the age of the smart artificial

intelligence: AI’s dual capacities for automating and

informating work. Business Information Review, 36(4),

178–187. Scopus. https://doi.org/10.1177/02663821

19883999

Navigating Responsible AI Adoption

343

Jarrahi, M. H., Askay, D., Eshraghi, A., & Smith, P. (2022).

Artificial intelligence and knowledge management: A

partnership between human and AI. Business Horizons.

https://doi.org/10.1016/j.bushor.2022.03.002

Jobin, A., Ienca, M., & Vayena, E. (2019). Artificial

Intelligence: The global landscape of ethics guidelines.

arXiv.Org. https://doi.org/10.48550/arXiv.1906.11668

Lehoux, P., Rivard, L., de Oliveira, R. R., Mörch, C. M., &

Alami, H. (2023). Tools to foster responsibility in

digital solutions that operate with or without artificial

intelligence: A scoping review for health and

innovation policymakers. International Journal of

Medical Informatics, 170, 104933. https://doi.org/

10.1016/j.ijmedinf.2022.104933

Leijnen, S., Aldewereld, H., van Belkom, R., Bijvank, R.,

& Ossewaarde, R. (2020). An agile framework for

trustworthy AI. NeHuAI@ ECAI, 75–78.

Lukkien, D. R. M., Nap, H. H., Buimer, H. P., Peine, A.,

Boon, W. P. C., Ket, J. C. F., Minkman, M. M. N., &

Moors, E. H. M. (2023). Toward Responsible Artificial

Intelligence in Long-Term Care: A Scoping Review on

Practical Approaches. Gerontologist, 63(1), 155–168.

Scopus. https://doi.org/10.1093/geront/gnab180

Mattu, J. A., Jeff Larson,Lauren Kirchner,Surya. (2016).

Machine Bias. ProPublica. https://www.propublica.

org/article/machine-bias-risk-assessments-in-criminal-

sentencing

Microsoft Corporation. (2022). Microsoft Responsible AI

Standard. https://blogs.microsoft.com/wp-content/

uploads/prod/sites/5/2022/06/Microsoft-Responsible-

AI-Standard-v2-General-Requirements-3.pdf

Minkkinen, M., Zimmer, M. P., & Mäntymäki, M. (2023).

Co-Shaping an Ecosystem for Responsible AI: Five

Types of Expectation Work in Response to a

Technological Frame. Information Systems Frontiers,

25(1), 103–121. Scopus. https://doi.org/10.1007/s

10796-022-10269-2

Mitchell, M., Wu, S., Zaldivar, A., Barnes, P., Vasserman,

L., Hutchinson, B., Spitzer, E., Raji, I. D., & Gebru, T.

(2019). Model Cards for Model Reporting. Proceedings

of the Conference on Fairness, Accountability, and

Transparency, 220–229. https://doi.org/10.1145/

3287560.3287596

Morley, J., Floridi, L., Kinsey, L., & Elhalal, A. (2020).

From What to How: An Initial Review of Publicly

Available AI Ethics Tools, Methods and Research to

Translate Principles into Practices. Science and

Engineering Ethics, 26(4), 2141–2168. Scopus.

https://doi.org/10.1007/s11948-019-00165-5

Murray, D. (1997). Ethics in Organizations. Kogan Page

Publishers.

Nagbøl, P. R., Müller, O., & Krancher, O. (2021).

Designing a Risk Assessment Tool for Artificial

Intelligence Systems: Vol. 12807 LNCS (p. 339).

Scopus. https://doi.org/10.1007/978-3-030-82405-

1_32

Narayanan, M., & Schoeberl, C. (2023). A Matrix for

Selecting Responsible AI Frameworks. Center for

Security and Emerging Technology. https://doi.

org/10.51593/20220029

Noiseau, P. (2023). Ethics of care and Artificial

Intelligence: The need to integrate a feminist normative

approach. In B. Prud’homme, C. Régis, G. Farnadi, V.

Dreier, S. Rubel, & C. d’Oultremont (Eds.), Missing

links in AI governance (pp. 344–358). Paris : UNESCO;

Montréal : Mila – Québec Institute of Artificial

Intelligence.

Obermeyer, Z., Nissan, R., Stern, M., Eaneff, S.,

Bembeneck, E. J., & Mullainathan, S. (2021).

Algorithmic Bias Playbook. Chicago Booth.

https://www.chicagobooth.edu/research/center-for-app

lied-artificial-intelligence/research/algorithmic-bias/pl

aybook

Oliveira, D., & Dalkir, K. (2022). Knowledge Capture for

the Design of a Technology Assessment Tool. 14th

International Joint Conference on Knowledge

Discovery, Knowledge Engineering and Knowledge

Management, 2, 185–192. https://doi.org/10.

5220/0011551400003335

Perry, N., Srivastava, M., Kumar, D., & Boneh, D. (2022).

Do Users Write More Insecure Code with AI

Assistants? (arXiv:2211.03622). arXiv. https://doi.org/

10.48550/arXiv.2211.03622

Probasco, E. (2022). A Common Language for Responsible

AI. Center for Security and Emerging Technology.

Qiang, V., Rhim, J., & Moon, A. (2023). No such thing as

one-size-fits-all in AI ethics frameworks: A

comparative case study. AI & Society. https://doi.

org/10.1007/s00146-023-01653-w

Reich, B. H., & Benbasat, I. (2000). Factors That Influence

the Social Dimension of Alignment between Business

and Information Technology Objectives. MIS

Quarterly, 24(1), 81–113.

Richards, J., Piorkowski, D., Hind, M., Houde, S., &

Mojsilović, A. (2020). A Methodology for Creating AI

FactSheets. http://arxiv.org/abs/2006.13796

Rolls Royce. (2021). The Aletheia Framework 2.0.

https://www.rolls-royce.com/~/media/Files/R/RollsRo

yce/documents/stand-alone-pages/aletheia-framework-

booklet-2021.pdf

Siala, H., & Wang, Y. (2022). SHIFTing artificial

intelligence to be responsible in healthcare: A

systematic review. Social Science and Medicine, 296.

Scopus. https://doi.org/10.1016/j.socscimed.2022.

114782

Skitka, L. J., Mosier, K. L., Burdick, M., & Rosenblatt, B.

(2000). Automation bias and errors: Are crews better

than individuals? International Journal of Aviation

Psychology, 10(1), 85–97. Scopus. https://doi.org/

10.1207/S15327108IJAP1001_5

Suresh, H., & Guttag, J. (2021). A Framework for

Understanding Sources of Harm throughout the

Machine Learning Life Cycle. ACM International

Conference Proceeding Series. Scopus. https://doi.

org/10.1145/3465416.3483305

Tursunbayeva, A., & Renkema, M. (2022). Artificial

intelligence in health-care: Implications for the job

design of healthcare professionals. Asia Pacific Journal

of Human Resources. https://doi.org/10.1111/1744-

7941.12325

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

344

Université de Montréal. (2018). Declaration of Montréal

for a responsible development of AI. https://

www.montrealdeclaration-responsibleai.com

Vassileva, B. K. (2021). Artificial Intelligence: Concepts

and Notions. In B. Vassileva & M. Zwilling (Eds.),

Advances in Human and Social Aspects of Technology

(pp. 1–18). IGI Global. https://doi.org/10.4018/978-1-

7998-4285-9.ch001

Verbeek, P.-P. & Tijink, Daniël. (2020). Guidance ethics

approach: An ethical dialogue about technology with

perspective on actions.

von Richthofen, G., Ogolla, S., & Send, H. (2022).

Adopting AI in the Context of Knowledge Work:

Empirical Insights from German Organizations.

Information, 13(4). https://doi.org/10.3390/

info13040199

World Economic Forum. (n.d.). Empowering AI

Leadership: An Oversight Toolkit for Boards of

Directors. World Economic Forum. https://

express.adobe.com/page/RsXNkZANwMLEf/.

Navigating Responsible AI Adoption

345