When we think about traditional technology services, such as the Internet and popular smartphone apps, interaction or experience related words tend to come to our mind. However, when we shift to more disruptive and breakthrough technologies, such as big data, IoT or Artificial Intelligence (AI), trust and security become a key challenge as the environments in which the personal information is stored and used to extract value changes in ways that are difficult to understand for normal users. This study explores the current challenges to trust and rely on a Virtual Assistant and establishes a series of constructs to build trust in a Telecoms Data Driven Virtual Assistant by considering aspects from an embryonic idea of the AI to its final articulation as a digital service. It examines a series of multi-methodology research interactions to help shape and refine the service, not only to build trust in the AI but also in the organisation behind it.