A cross-disciplinary contribution focussed on how through a combination of research practices including causal layered legal analysis and speculative design workshops, we can improve the lack of anticipation and multi-stakeholder participation in the governance of AI. Despite the many initiatives towards regulating emerging technologies such as AI, current policies and regulations are insufficient in dealing with the many complex challenges and harmful disruptions for people and societies around the world. To obtain the deeper insights into the causes for our lack of transformative actions to ensure trusted and trustworthy AI, it is argued we need not only innovative governance but to innovate research to provide us with the deeper insights we need to understand not only the problems and causes. Through a combination of Speculative AI Futures workshops and Causal Layered Analysis we can improve recommendations considering how our worldviews, myths and metaphors hinder or help us to take action towards any of our co-created desirable futures.