Any decision about using artificial intelligence in monitoring and evaluation involves evidence needs, workflow, and capability considerations. In this blog post, Douglas Glandon explains why we need to consider all these dimensions together, not separately, to make meaningful decisions on AI use.