Quality Attributes

Originating from software engineering, quality attributes are measurable non-functional requirements that define how a system operates. The international standard ISO/IEC 25010 (SQuaRE) provides a comprehensive model, categorizing them into eight characteristics: functional suitability, performance efficiency, compatibility, usability, reliability, security, maintainability, and portability. In enterprise risk management, these attributes serve as benchmarks for assessing operational and compliance risks. For AI systems, this concept is extended by frameworks like the NIST AI Risk Management Framework to include trustworthiness characteristics such as fairness, explainability, and robustness, crucial for managing the unique ethical and societal risks associated with AI.

Practical application involves a three-step process. First, Define and Prioritize: Based on business context and regulatory obligations like GDPR, enterprises select critical quality attributes from frameworks like ISO/IEC 25010 and establish quantifiable acceptance criteria, such as 99.9% system availability for 'reliability.' Second, Integrate into Lifecycle: These requirements are embedded into the system development lifecycle (SDLC), including implementing Privacy by Design principles and conducting bias audits. Third, Monitor and Govern: Post-deployment, automated dashboards track performance against quality metrics, with regular audits ensuring ongoing compliance. This approach can measurably reduce risk events and improve audit pass rates.

Taiwan enterprises face three primary challenges. First, Resource and Skill Gaps: Many SMEs lack dedicated SQA or AI ethics teams. Second, Ambiguous Metrics: For emerging AI attributes like fairness, standardized benchmarks are often missing. Third, Inadequate Data Governance: Pre-existing data biases undermine an AI model's reliability and fairness. To overcome these, firms should adopt international frameworks like the NIST AI RMF, leverage open-source tools for automated testing, and establish cross-functional AI governance teams to create a unified, internal metrics library. Prioritizing data governance is the critical first step.

Winners Consulting specializes in quality attributes for Taiwan enterprises, delivering compliant management systems within 90 days. Free consultation: https://winners.com.tw/contact