With the wide scale adoption of cloud computing paradigms and Internet of Things implementations, the arena of service delivery has changed significantly for the world of technology-oriented services. While these new modes of service delivery have provided innovations that were not fathomable earlier, they not only create new non-traditional pathways for evidence generation, nevertheless, they have also been seen to carry a potential of creating and exposing a whole new set of security vulnerabilities that malicious actors can exploit. Traditional forensic investigation process falls short in many ways including the way it aims to identify and link crucial pieces of evidence in these scenarios due to a myriad of internal and external factors. Challenges and inadequacies have been identified in the community in relation to these specific sets of advancing technologies that hinder the efficient identification, extraction and processing of evidence generated in crimes involving them. In this paper we wish to surmise the current forensic collection methods, how they pose challenges when put together with the latest cloud and IoT technologies and then attempt to propose an advanced, more intuitive, and theoretically validated digital forensic collection process guided by zero trust principle and enforced with artificial intelligence and machine learning methodologies with the aim to fill the gaps and provide a more reliable, robust and intuitive model to perform digital evidence collection in the given context.