Module D: Selection of Practical Measures
ModulesModule D: Selection of Practical Measures
Module D introduces the AIOLIA Portfolio of Measures as the central operational tool for translating high-level AI ethics requirements into concrete technical and organizational interventions. Building directly on the issues identified in Module C, it provides a structured mechanism to move from ethical risk analysis to practical implementation.
The module introduces a structured 4-step selection and refinement process that guides learners through navigating the 122 measures contained within the portfolio. Rather than treating the portfolio as a rigid compliance checklist, Module D emphasizes contextual judgment, ethical reflection, and trade-off management—training learners to select, adapt, prioritize, and streamline measures to ensure they address all critical ethics issues while remaining technically and organizationally feasible.
Learning Goals
- Navigate and apply the AIOLIA Portfolio of Measures (comprising 122 technical, organizational, and dual-nature measures across the 7 ALTAI requirements).
- Apply the 4-step process to prioritize identified requirements, filter relevant measures, adapt entries to specific contexts, and streamline final selections.
- Distinguish between Technical (TECH), Organizational (ORG), and Combined (BOTH) measures, ensuring balanced governance and design safeguards.
- Fully operationalize selected measures by defining core components: relevance, implementation steps, verification criteria, potential challenges, risks, and responsible roles.
- Recognize ethical urgency vs. technical feasibility to avoid "measure overkill" and prepare selected interventions for tension analysis in Module E.
What is this about?
Why is this important?
Selection of Practical Measures
The transition from ethics issue identification to concrete system safeguards relies on the AIOLIA Portfolio of Measures, a decision-support repository of 122 technical (TECH), organizational (ORG), and combined (BOTH) interventions organized across the seven ALTAI requirements. To prevent mechanical compliance, learners apply a structured 4-step process: first, prioritizing the ALTAI requirements identified during evaluation; second, targeted browsing of portfolio sections corresponding to those high-priority requirements; third, selecting, combining, and adapting measures to fit the specific operational context; and fourth, reviewing the set for ethical urgency and technical feasibility to maintain a streamlined, non-redundant list.
A central core of the module is transforming one-sentence measure summaries into fully specified, operational ethics interventions. A complete measure must go beyond a description to define why it is relevant, how it will be technically or organizationally achieved, how fulfillment will be assessed or audited, what operational challenges or risks exist if neglected, and which specific stakeholder roles (e.g., QA teams, ML engineers, compliance officers) hold primary accountability.
Finally, the module highlights that technical safeguards and organizational governance are inherently interdependent. A technically robust system operating in a poorly governed organization remains an ethical risk, just as strong policy controls cannot rescue a flawed technical architecture. By teaching learners to balance design controls (like confidence scores or oversight overrides) with organizational protocols (like mandatory human sign-offs and bias audits), Module D prepares the selected measures for real-world deployment and downstream tension analysis.
