A Method for Generating Dynamic Responsible AI Guidelines for Collaborative Action

by   Marios Constantinides, et al.

To improve the development of responsible AI systems, developers are increasingly utilizing tools such as checklists or guideline cards to ensure fairness, transparency, and sustainability. However, these tools face two main challenges. First, they are static and are not meant to keep pace with the latest responsible AI literature and international standards. Second, they tend to prioritize individual usage over fostering collaboration among AI practitioners. To overcome these limitations, we propose a method that enables easy updates of responsible AI guidelines by incorporating research papers and ISO standards, ensuring that the content remains relevant and up to date, while emphasizing actionable guidelines that can be implemented by a wide range of AI practitioners. We validated our method in a case study at a large tech company by designing and deploying a tool that recommends interactive and actionable guidelines, which were generated by a team of engineers, standardization experts, and a lawyer using our method. Through the study involving AI developers and engineers, we assessed the usability and effectiveness of the tool, showing that the guidelines were considered practical and actionable. The guidelines encouraged self-reflection and facilitated a better understanding of the ethical considerations of AI during the early stages of development, significantly contributing to the idea of "Responsible AI by Design" – a design-first approach that considers responsible AI values throughout the development lifecycle and across business roles.


page 6

page 11

page 13

page 16


Software Engineering for Responsible AI: An Empirical Study and Operationalised Patterns

Although artificial intelligence (AI) is solving real-world challenges a...

Dislocated Accountabilities in the AI Supply Chain: Modularity and Developers' Notions of Responsibility

Responsible AI guidelines often ask engineers to consider how their syst...

Model Reporting for Certifiable AI: A Proposal from Merging EU Regulation into AI Development

Despite large progress in Explainable and Safe AI, practitioners suffer ...

A Comparative Analysis of Industry Human-AI Interaction Guidelines

With the recent release of AI interaction guidelines from Apple, Google,...

Behavioral Use Licensing for Responsible AI

Scientific research and development relies on the sharing of ideas and a...

Training Ethically Responsible AI Researchers: a Case Study

Ethical oversight of AI research is beset by a number of problems. There...

How AI Developers Overcome Communication Challenges in a Multidisciplinary Team: A Case Study

The development of AI applications is a multidisciplinary effort, involv...

Please sign up or login with your details

Forgot password? Click here to reset