13th Conference on Cloud Computing Conference, Big Data & Emerging Topics

Ensuring Quality in the OECD AI Lifecycle Through

ISO/IEC Standards
Juan Ignacio Torres [0000-0002-9399-7561], Ariel Pasini [0000-0002-4752-7112],
Patricia Pesado [0000-0003-0000-3482]
Institute of Research in Computer Science LIDI (III-LIDI), Faculty of Computer Science,
National University of La Plata, Argentina
{jitorres,apasini,ppesado}@lidi.info.unlp.edu.ar

Abstract. The integration of artificial intelligence (AI) technologies within or-

ganizations presents both significant opportunities and complex challenges. To
manage this complexity, ISO/IEC standards provide a structured framework for
the adoption and management of AI systems throughout their lifecycle. This ar-
ticle explores the role of ISO/IEC standards in ensuring the quality, security, and
ethical alignment of AI systems, based on the lifecycle framework defined by the
Organisation for Economic Co-operation and Development (OECD). The paper
outlines how these standards support AI system development, from planning and
design through deployment and monitoring, addressing critical issues such as
governance, data quality, bias detection, and system reliability. A comprehensive
quality model is proposed, drawing on ISO/IEC 25058 and 25059 standards, to
assess the effectiveness and transparency of AI systems in real-world environ-
ments. The adoption of these standards is shown to enhance corporate reputation,
improve regulatory compliance, and mitigate risks, positioning organizations to
leverage AI technologies responsibly and efficiently.

Keywords: AI Governance, ISO/IEC Standards, AI System Lifecycle, Quality

Management

1 Introduction

The adoption of artificial intelligence technologies simultaneously represents both an

opportunity and a challenge for contemporary organizations. In an increasingly com-
petitive and dynamic industrial environment, AI has become a key driver of innovation,
operational efficiency, and the personalization of products and services. Its revolution-
ary capabilities range from process automation to the optimization of strategic decision-
making, impacting sectors such as healthcare, finance, and logistics [1,2].
However, this advancement comes with technical complexities, ethical considera-
tions, and operational challenges that may lead to issues without a structured imple-
mentation framework. The ISO/IEC standards related to AI emerge as a structured re-
sponse to this need, providing organizations with methodological guidelines, best prac-
tices, and conceptual frameworks that facilitate the effective integration of intelligent
systems into their operations and strategies. These standards not only clarify technical
aspects but also provide tools for the organizational management of these technologies,
helping companies maximize the benefits of AI [3].

- 90 -
 13th Conference on Cloud Computing Conference, Big Data & Emerging Topics

In this context, the present article aims to survey quality standards applicable to the
lifecycle of an AI system, according to the framework defined by the OECD [4]. Addi-
tionally, a comprehensive quality model for AI software will be proposed to establish
specific criteria for measuring and ensuring the effectiveness, security, and transpar-
ency of these systems in real-world environments.

2 General Concepts

2.1 Quality Standards

ISO quality standards are international standards developed by the International Organ-
ization for Standardization (ISO) that establish guidelines and requirements to ensure
that products, services, and systems meet globally recognized criteria for quality,
safety, and efficiency. These standards result from international agreements among ex-
perts and cover a wide range of activities, from product manufacturing to process man-
agement and service delivery [5].
The adoption of these standards enables organizations to optimize their internal pro-
cesses, reduce errors, and enhance customer satisfaction. By adhering to internationally
recognized criteria, companies can improve their reputation and gain access to new
markets, demonstrating their commitment to quality and excellence in their operations
[6].

2.2 Artificial Intelligence

Artificial intelligence (AI) is a field of computer science focused on creating systems
capable of performing tasks that typically require human intelligence, such as learning,
reasoning, and perception. These systems can analyze data, recognize patterns, and
make decisions with a certain degree of autonomy [7].
Artificial intelligence is based on algorithms and mathematical models that enable
machines to learn from experience and adapt to new information inputs. By leveraging
these technologies, computers can be trained to perform specific tasks by analyzing
large volumes of data and identifying patterns within them. This has led to the devel-
opment of applications across various sectors, transforming how we interact with tech-
nology in our daily lives [8,9].

3 Quality Standards in Artificial Intelligence

The ISO/IEC standards for AI provide organizations with a safer pathway for adopting
intelligent technologies, reducing the uncertainty and risks inherent in the development
of complex systems. Through proven methodological guidelines, these standards help
identify and mitigate potential technical issues before full deployment, ensuring com-
pliance with emerging regulations and minimizing the risk of legal and operational
complications.

- 91 -
 13th Conference on Cloud Computing Conference, Big Data & Emerging Topics

Implementing these standards optimizes the investment of technological and finan-

cial resources by establishing standardized processes that accelerate AI project devel-
opment with a lower probability of errors. Organizations achieve better interoperability
and scalability of their systems, promoting flexible designs that facilitate integration
between different technologies and adaptation to changing needs while fostering the
reuse of knowledge and experiences across departments.
Additionally, ISO/IEC standards strengthen corporate trust and improve relation-
ships with stakeholders by providing transparency frameworks for the use of algorithms
and data. Organizations can demonstrate their commitment to ethical and responsible
AI practices, enhancing their reputation among customers, regulatory authorities, and
investors while simultaneously transforming their internal operations with reliable and
adaptable intelligent systems [10].

4 Lifecycle of AI Systems according to the OECD

The Organisation for Economic Co-operation and Development has defined an AI sys-
tem lifecycle that consists of several key phases (Figure 1). This cycle begins with plan-
ning and design, where the system’s objectives and requirements are established, con-
sidering the specific context and needs. Next, the data collection and processing phase
ensures that the gathered information is relevant and of high quality for model training.
Following this, the model development phase involves creating or adapting algorithms
to perform the defined tasks.
Once the model is built, it enters the verification and validation phase, where its
performance is assessed, and adjustments are made to ensure effectiveness and security.
After passing these tests, the system moves to the deployment phase for operational
use. During the operation and monitoring phase, the system’s performance is continu-
ously supervised, identifying potential improvements or necessary modifications.
It is important to note that these phases are not strictly sequential and may be itera-
tive, depending on project needs. Furthermore, an AI system may be decommissioned
at any time during the operation and monitoring phase, particularly if it fails to meet its
objectives or poses unacceptable risks.

50
40
Data A
30
20
10
0
0 5 10 15 20 25 30

Fig. 1. AI Model Development Lifecycle according to OECD.

- 92 -
 13th Conference on Cloud Computing Conference, Big Data & Emerging Topics

5 Application of Quality Standards in the OECD Lifecycle

The lifecycle of artificial intelligence systems, defined by the OECD, comprises a set
of interconnected stages that require a specific regulatory framework. Each phase of
intelligent system development has ISO/IEC international standards that provide pre-
cise methodological guidelines and implementation criteria.
In the initial planning and design phase, ISO/IEC 38507:2022 (Information Tech-
nology - Governance of IT — Governance implications of the use of artificial intelli-
gence by organizations) and ISO/IEC 42001:2023 (Information Technology - Artificial
intelligence — Management system) standards establish guidelines for strategic gov-
ernance and objective definition, facilitating a structured framework for the conceptu-
alization of artificial intelligence systems. During data collection and processing,
ISO/IEC 5259 (Artificial intelligence — Data quality for analytics and machine learn-
ing (ML)) standards offer detailed guidance for data quality management, while
ISO/IEC TR 24027:2021 (Information Technology - Artificial intelligence (AI) — Bias
in AI systems and AI aided decision making) provides tools for bias detection and mit-
igation in data sets.
The construction and verification of the model are supported by standards such as
ISO/IEC 23053 (Framework for Artificial Intelligence Systems Using Machine Learn-
ing), which outlines the architecture of machine learning systems, and ISO/IEC
25059:2023 (Software Engineering — Systems and Software Quality Requirements
and Evaluation (SQuaRE) — Quality Model for AI Systems), which defines quality
criteria for designing AI systems.
During the deployment and operational phases, ISO/IEC 5338:2023 (Information
Technology — Artificial Intelligence — AI System Life Cycle Processes) and ISO/IEC
TR 5469:2024 (Artificial Intelligence — Functional Safety and AI Systems) establish
processes for operational transition and functional safety assurance, enabling continu-
ous monitoring and the ability to decommission the system if necessary.
To evaluate the quality of artificial intelligence systems throughout these stages, an
assessment model is proposed based on ISO/IEC 25058 (Systems and software engi-
neering — Systems and software Quality Requirements and Evaluation (SQuaRE) —
Guidance for quality evaluation of artificial intelligence (AI) systems) and 25059 stand-
ards, aligned with the OECD framework.
In the planning and design phase, the evaluation will focus on the proper application
of governance principles, verifying the system’s alignment with strategic objectives and
risk identification, as established by ISO/IEC 38507 and 42001.
During the data collection and processing phase, the quality and representativeness
of the data used in system development will be analyzed. By applying the criteria from
ISO/IEC 5259 and ISO/IEC TR 24027:2021, the accuracy, completeness, and fairness
of the data will be measured, with a focus on bias detection and mitigation.
In the model construction and verification phase, the robustness and reliability of the
AI system will be assessed, considering the quality criteria defined in ISO/IEC 25059.
The system's architecture will be analyzed in accordance with ISO/IEC 23053, meas-
uring its generalization capability, model explainability, and suitability for the defined
purpose.

- 93 -
 13th Conference on Cloud Computing Conference, Big Data & Emerging Topics

For the deployment and operation phase, the evaluation will focus on the security
and stability of the system in real-world environments. The ISO/IEC 5338 and TR 5469
standards will be applied to ensure the presence of protection mechanisms against vul-
nerabilities and failure mitigation procedures. In addition, ISO/IEC 25059 will be used
to assess the system’s quality in terms of reliability and operational risk control.
Finally, in the monitoring and adjustment phase, the evaluation will continue through
periodic audits and model degradation metrics. Strategies defined in ISO/IEC 25058
will be implemented to ensure the maintainability and evolution of the system, guaran-
teeing its reliability over time.

6 Conclusions

The adoption of AI in organizations represents a complex process that requires a sys-

tematic and structured approach. ISO/IEC standards emerge as a fundamental tool for
managing this complexity, providing a regulatory framework that covers technical as-
pects while also addressing strategic, ethical, and operational dimensions.
The benefits of these standards go beyond the technical, becoming a competitive
advantage that enhances corporate reputation, facilitates regulatory compliance, and
optimizes investment in smart technologies. They reduce risks, establish standardized
processes, promote interoperability, and demonstrate a commitment to ethical and
transparent practices.
Throughout the article, a comprehensive model was proposed to analyze the quality
of AI systems at each phase of their lifecycle. Based on ISO/IEC 25058 and 25059
standards, the model allows for a detailed assessment that spans from initial planning
to continuous monitoring, ensuring risk identification, bias mitigation, and verification
of the robustness of intelligent systems.

References
1. Alhosani, K., Alhashmi, S.M.: Opportunities, challenges, and benefits of AI innovation in
government services: a review. Discov Artif Intell 4, 18 (2024).
https://doi.org/10.1007/s44163-024-00111-w
2. Ali, O., Abdelbaki, W., Shrestha, A., Elbasi, E., Alryalat, M.A.A., Dwivedi, Y.K.: A sys-
tematic literature review of artificial intelligence in the healthcare sector: Benefits, chal-
lenges, methodologies, and functionalities. J. Innov. Knowl. 8(1), 100333 (2023).
https://doi.org/10.1016/j.jik.2023.100333
3. Oviedo, J., Rodriguez, M., Trenta, A., Cannas, D., Natale, D., Piattini, M.: ISO/IEC quality
standards for AI engineering. Comput. Sci. Rev. 54, 100681 (2024).
https://doi.org/10.1016/j.cosrev.2024.100681
4. OECD Council: Recommendation of the Council on Artificial Intelligence. https://le-
galinstruments.oecd.org/en/instruments/OECD-LEGAL-0449. Last accessed 2025/03/20.
5. Bruijn, H., Duin, R., Huijbregts, M. A., Guinee, J. B., Gorree, M., Heijungs, R., Huppes, G.,
Kleijn, R., Koning, A., Oers, L., Sleeswijk, A. W.: Handbook on life cycle assessment: op-
erational guide to the ISO standards. Kluwer Academic Publishers, Dordrecht, The Nether-
lands (2004).

- 94 -
 13th Conference on Cloud Computing Conference, Big Data & Emerging Topics

6. Su, H.-C., Dhanorkar, S., Linderman, K.: A competitive advantage from the implementation
timing of ISO management standards. Journal of Operations Management, 37, 31–44
(2015). https://doi.org/10.1016/j.jom.2015.03.004
7. Abbass, H.: Editorial: What is Artificial Intelligence? IEEE Transactions on Artificial Intel-
ligence, 2(2), 94–95 (2021). https://doi.org/10.1109/TAI.2021.3096243
8. Makridakis, S.: The forthcoming Artificial Intelligence (AI) revolution: Its impact on society
and firms. Futures, 90, 46–60 (2017). https://doi.org/10.1016/j.futures.2017.03.006
9. Pedro, F., Subosa, M., Rivas, A., & Valverde, P.: Artificial intelligence in education: Chal-
lenges and opportunities for sustainable development. (2019).
10. Zielke, T.: Is Artificial Intelligence Ready for Standardization? In: Yilmaz, M., Niemann,
J., Clarke, P., Messnarz, R. (eds.) Systems, Software and Services Process Improvement.
EuroSPI 2020. Communications in Computer and Information Science, vol. 1251, Springer,
Cham (2020). https://doi.org/10.1007/978-3-030-56441-4_19

- 95 -