2019/11/27 by Ehsan Toreini, Toreini, Ehsan, Mhairi Aitken +9 · 6 citations
Computer Science · Social Sciences · #Adversarial Robustness in Machine Learning #Artificial Intelligence (cs.AI) #Computers and Society (cs.CY) #Ethics and Social Impacts of AI #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Machine Learning (cs.LG)
paper · pdf · doi:10.48550/arxiv.1912.00782
openalex publication_date 2019/11/27 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
To build AI-based systems that users and the public can justifiably trust one\nneeds to understand how machine learning technologies impact trust put in these\nservices. To guide technology developments, this paper provides a systematic\napproach to relate social science concepts of trust with the technologies used\nin AI-based services and products. We conceive trust as discussed in the ABI\n(Ability, Benevolence, Integrity) framework and use a recently proposed mapping\nof ABI on qualities of technologies. We consider four categories of machine\nlearning technologies, namely these for Fairness, Explainability, Auditability\nand Safety (FEAS) and discuss if and how these possess the required qualities.\nTrust can be impacted throughout the life cycle of AI-based systems, and we\nintroduce the concept of Chain of Trust to discuss technological needs for\ntrust in different stages of the life cycle. FEAS has obvious relations with\nknown frameworks and therefore we relate FEAS to a variety of international\nPrincipled AI policy and technology frameworks that have emerged in recent\nyears.\n