EthicsAccountabilityTransparencycs.AI
Software Engineering for Responsible AI: An Empirical Study and Operationalised Patterns
Qinghua Lu, Liming Zhu, Xiwei Xu, Jon Whittle, David Douglas, Conrad Sanderson
November 18, 2021
2022 IEEE/ACM 44th International Conference on Software Engineering: Software Engineering in Practice (ICSE-SEIP)
53 citations
5 influential
Key Takeaways
-
AI ethics principles exist but lack practical guidance: The study found that while governments and organizations have published numerous AI ethics principles and guidelines, these remain "high-level" and don't provide concrete steps for developers to actually implement them in real systems—a gap that practitioners struggle to bridge.
-
Practitioners need actionable patterns, not just principles: Through interviews with 21 AI scientists and engineers, the research identified that developers require specific, operationalized patterns and templates to translate abstract ethics concepts into actual design and development decisions for responsible AI systems.
-
A new template can make ethics principles concrete: The authors propose a template that converts high-level AI ethics principles into practical patterns, offering developers a structured way to implement responsible AI practices rather than relying on interpretation of vague guidelines.
-
The gap between ethics theory and practice is significant: The empirical study reveals a disconnect between what ethics guidelines recommend and what practitioners can realistically implement, suggesting that the AI industry needs better tools and frameworks to operationalize responsible AI development.
Abstract
Although artificial intelligence (AI) is solving real-world challenges and transforming industries, there are serious concerns about its ability to behave and make decisions in a responsible way. Many AI ethics principles and guidelines for responsible AI have been recently issued by governments, organisations, and enterprises. However, these AI ethics principles and guidelines are typically high-level and do not provide concrete guidance on how to design and develop responsible AI systems. To address this shortcoming, we first present an empirical study where we interviewed 21 scientists and engineers to understand the practitioners' perceptions on AI ethics principles and their implementation. We then propose a template that enables AI ethics principles to be operationalised in the form of concrete patterns and suggest a list of patterns using the newly created template. These patterns provide concrete, operationalised guidance that facilitate the development of responsible AI systems.