| Inter-Pillar Relationships |
| Pillar: Operational Integrity |
|
Governance vs. Robustness |
■ Reinforcing
|
Governance frameworks support robustness by establishing guidelines to ensure AI systems are resilient and reliable
(Batool et al., 2023 )
|
AI governance mandates robustness assurance in healthcare to ensure reliable system performance under varying conditions
(Bullock et al., 2024 )
|
|
Interpretability vs. Robustness |
■ Tensioned
|
Interpretability can compromise robustness due to increased complexity in models
(Hamon et al., 2020 )
|
Interpretable models in safety-critical applications may reduce robustness, increasing vulnerability to adversarial attacks
(Hamon et al., 2020 )
|
|
Explainability vs. Robustness |
■ Tensioned
|
High explainability may simplify models, potentially reducing their robustness
(Rudin, 2019 )
|
Simplified credit scoring models for explainability may perform poorly under non-standard conditions
(Rudin, 2019 )
|
|
Robustness vs. Security |
■ Reinforcing
|
Robustness strengthens security against adversarial attacks, enhancing overall system reliability
(Habbal et al., 2024 )
|
Robust AI enhances security by withstanding data poisoning, crucial in cybersecurity
(Habbal et al., 2024 )
|
|
Robustness vs. Safety |
■ Reinforcing
|
Robustness ensures AI operates safely under challenging conditions, enhancing overall safety
(Leslie, 2019 )
|
In medicine, robust AI systems reliably identify anomalies despite data distribution changes, promoting safety
(Leslie, 2019 )
|
| Cross-Pillar Relationships |
| Pillar: Ethical Safeguards vs. Operational Integrity |
|
Fairness vs. Robustness |
■ Tensioned
|
Fairness might necessitate modifications that decrease robustness
(Tocchetti et al., 2022 )
|
Adjustments to AI models for fairness in loan approvals might reduce performance across datasets
(Braiek & Khomh, 2024 )
|
|
Inclusiveness vs. Robustness |
■ Tensioned
|
Inclusiveness may compromise robustness by prioritizing accessibility over resilience in challenging scenarios
(Leslie, 2019 )
|
AI built for varied groups might underperform in rigorous environments, highlighting tension
(Leslie, 2019 )
|
|
Bias Mitigation vs. Robustness |
■ Reinforcing
|
Bias mitigation enhances robustness by incorporating diverse data, reducing systematic vulnerabilities in AI models
(Ferrara, 2024 )
|
Inclusive datasets in AI model training improve both bias mitigation and system robustness
(Ferrara, 2024 )
|
|
Accountability vs. Robustness |
■ Tensioned
|
Accountability’s demand for traceability may compromise robustness by exposing sensitive vulnerabilities
(Leslie, 2019 )
|
In autonomous vehicles, robust privacy measures can conflict with traceability required for accountability
(Busuioc, 2021 )
|
|
Privacy vs. Robustness |
■ Tensioned
|
Achieving high privacy can sometimes challenge robustness by limiting data availability
(Hamon et al., 2020 )
|
Differential privacy techniques may decrease robustness, impacting AI model performance in varied conditions
(Hamon et al., 2020 )
|
| Pillar: Operational Integrity vs. Societal Empowerment |
|
Robustness vs. Sustainability |
■ Tensioned
|
Minimizing energy consumption could compromise robustness under variable conditions
(Carayannis & Grigoroudis, 2023 )
|
Energy-efficient machine learning models may struggle with edge-case data
(Braiek & Khomh, 2024 )
|
|
Human Oversight vs. Robustness |
■ Reinforcing
|
Human oversight strengthens robustness by mitigating risks through active monitoring and intervention
(Tocchetti et al., 2022 )
|
Human oversight ensures robust system behavior during AI deployment in high-stakes environments like aviation
(High-Level Expert Group on Artificial Intelligence, 2020 )
|
|
Robustness vs. Transparency |
■ Reinforcing
|
Robustness enhances transparency by providing consistent operations, reducing opaque behaviors
(Hamon et al., 2020 )
|
Greater AI robustness minimizes erratic outcomes, facilitating clearer system transparency
(Hamon et al., 2020 )
|
|
Robustness vs. Trustworthiness |
■ Reinforcing
|
Robustness directly contributes to the trustworthiness of AI by enhancing operational reliability under diverse conditions
(Braiek & Khomh, 2024 )
|
AI models with robust architectures improve trust by reliably handling environmental changes without function loss
(Braiek & Khomh, 2024 )
|
RAISEF
