Ethical Guidelines for the Deployment of Social and Behavior Change Chatbots
This resource was initially developed for Girl Effect’s internal staff involved in the design, development, and deployment of AI-powered chatbots for girls and young women in countries like South Africa, Kenya, and India, in order to provide them with guidance on the ethical dimensions of this complex field in the face of rapid technological developments.
It has since been adapted and made available publicly, as we hope it will prove useful to other teams grappling with best practice, whether they are new to chatbots/AI, or simply seeking to invest more heavily in ethical AI. The recommendations in this guidance have also been compiled as an implementation checklist, available on request.
This guidance is based on desk research accurate at the time of writing, and builds on the authors’ expertise in digital best practice across data rights and privacy, design, and digital safeguarding, as well as on the insights gathered during convenings with 1300+ global members of the Natural Language Processing Community of Practice. In addition, we also consulted 104 Girl Effect staff members, vendors, and Youth Advisors in Europe, Africa and India, to understand their level of understanding, attitudes, and concerns when it comes to the subject of ethical AI
You might also like
-
New brief: Artificial intelligence in the humanitarian sector
-
REvaluation week podcast episode: “Does AI really save work in evaluation?”
-
What’s next for the Gender, AI and MERL Working Group? An event series built in partnership with you!
-
Event: Gender, AI and MERL Working Group meeting focusing on gender inclusion and GenAI