The prevalence of an aging population and the associated increase in chronic disease and disability has led to policy makers recommending a shift in care delivery to self-management patient centric support models of care [1, 2]. Underpinning evidence, which signals a diminishing health workforce, is a rapid deployment in digital services and increasing recognition for deployment of Artificial Intelligence (AI), Machine Learning (ML) and Robotics to optimise service delivery [3, 4].  Implementation of such technologies to become mainstream in society raises a number of questions for targeted action.  Risk reduction particularly around consent or lack of control of data management, potential bias and discrimination are some examples of key issues, which policy administrators are rigoursly engaged on. Key principles underpinning the discussion include Trustful AI and Ethical AI, and there has been a significant number of European online workshops and discussions published on the topic on the past 12 months [5, 6]. Cross cutting themes to address these aforementioned principles include strong governance frameworks, which can provide guidance on regulation data privacy and security issues.  It is anticipated that AI will improve health systems performance in accuracy, productivity, and workflow, but what impact will it have on existing defined roles for health care professionals, and how many health care professional are actively engaged in debating and development of such systems determining how privacy, bias and security can be managed ethically [7,8]. 

Some recent policy actions instigated by the WHO include developing a standardised assessment framework for the evaluation of AI for health assessment diagnosis, and treatment [9, 10].  Within the domain of health informatics, the short to medium term goals are signposted to address data structuring, data quality and those related standards, which can act as a springboard for future AI adoption uptake and use in the health and social care. As healthcare systems become more complex integrated and connected, policy makers scan the Horizon and deliberate on how System of Systems (SoS) and Systems of Interest (SoI) will operate. Such  terms are new to health care professionals but briefly they relate to machine learning algorithms, which collect data, and which operate and interact with other systems in an operational environment which can combine the data from two or more SoI to provide a recommendation or result to the end user. Further information on this topic is available from Systems and software System of systems (SoS) considerations in life cycle stages of a system ISO/IEC/IEEE 21839:2019 [11,12]. If machine algorithms are making decisions based on the data in the SoI it is therefore critical that the data is of high quality and available. My experience in dealing with heterogeneous databanks to data would suggest that we have a long way to go before we can endorse such systems. 

​