Artificial Intelligence and Machine Learning are increasingly seen as key technologies for building more decentralised and resilient energy grids, but researchers must consider the ethical and social implications of their use Energy grids are undergoing rapid changes, requiring new ways both to process the large amounts of data generated from the power system, but also -increasingly -to take smart operational decisions [1].On the data side, the UK and most EU countries have committed to a target of offering a smart meter to every home by 2020 [2], with similar monitoring being installed in other parts of the energy network. This has led to some to refer to a "data tsunami", requiring development of new machine learning techniques to deal with the ensuing challenge of extracting useful information from this data -often in real time.Another trend is the use of AI techniques (such as those from multi-agent systems, computational game theory and decision making under uncertainty) to take autonomous allocation and control decisions. This is driven increasingly by the moves towards more decentralised energy systems, where prosumers (consumers with own micro-generation and storage) can generate and source their own electricity through peer-to-peer (P2P) trading in local energy markets and community energy schemes.To illustrate these trends from a UK perspective (with which the authors are most familiar), a number of projects are looking at computational techniques suitable for more decentralised energy future -including CESI (the UK National Centre for Energy Systems Integration) [23] and Responsive Flexibility, one of the largest smart energy demonstrators in the UK, focused on the islands of Orkney [26]. In fact, Community Energy Scotland (an organisation supporting such projects in Scotland) lists no less than 300 projects on their website [3]. This has attracted a surge of interest and considerable investments in technologies enabling these settings, such as blockchain, machine learning (ML) and distributed Artificial Intelligence (AI) [4].Yet, there are substantial ethical and social questions to be asked in such developments. The AI community has begun to recognise the challenges posed by rapid adoption of AI in the real world, and a number of general guidelines have been proposed [5]. Similarly, the smart grid community needs to consider the ethical challenges that the rapid adoption of AI and ML techniques to control our energy systems brings. In this piece, we discuss some domainspecific challenges and proposed solutions for adopting AI techniques to automate smart energy grids. We note our starting point is not purely generic AI ethical principles (although we have consulted recent reports, such as Floridi [5], which provide a useful guide). Rather, we consider the trends that have shaped AI adoption in smart grids in recent years, and illustrate the type of ethical challenges involved in each.