Introducing what a Humanitarian Code of Conduct for AI would look like. What are the principles we all agree on? What are the difficult areas? How do we ensure we make the code relevant to local implementers and communities?