Professor Tony Prescott co-authored the whitepaper on Robotics in Social Care: A Connected Care EcoSystem for Independent Living for the UK RAS Network. Here, we highlight the eight key ethical and societal issues identified in the whitepaper.
Security and Privacy
One of the key concerns for a connected system is data security. An assistive robot will need access to historic information, as well as information that is current to create a dynamic world view and adapt to the user’s changing needs and circumstances. Adapting behaviour and responsiveness requires drawing on information from environmental and activity sensors instrumented into a smart home, as well as information about the user’s current physical, cognitive and emotional state through vision, audition, touch, and physiological and biometric sensing. This is sensitive information that shared inappropriately could compromise an individual’s dignity and right to privacy. Maintaining independence needs to be balanced with ensuring a person’s autonomy and agency— we must design systems that people find trustworthy, not only from a reliability point of view, but also with respect to assurance of user privacy and security.
A related challenge is transparency of function. It should be clear to users, in general terms, what a RAS system does and what it is unable to do, why it is carrying out a task at a particular place or time, what data it is collecting for that purpose and whether this is being shared. The system needs to be able to give an account of itself in everyday terms that users understand and to respond promptly to requests to change behaviour.
Over time, people may form emotional bonds with RAS technologies, as they already do with devices such as mobile phones, tablets and cars. However, there may be a particularly strong tendency to develop ties with animate systems that have a social function such as companion robots. Careful consideration needs to be given to the design of these systems to ensure that the relationships people form with these technologies do not interfere with other aspects of their lives. For instance, for many purposes RAS technologies could resemble existing white goods, or seamlessly integrate with the fabric of the home, rather than having a life-like or humanoid form.
Autonomy of decision-making
As these systems become more complex they will be expected to make more critical decisions about the nature of the care they provide. While many circumstances in which RAS will be used in care are foreseeable, it is likely that future systems will sometimes need to reason about the situation in which they are operating and reach their own conclusions as to the best path of action. The design of systems that can plan effectively, whilst taking into account moral codes, is an active but relatively new area of research. A precautionary approach is required to ensure that machines do not operate outside the window of action for which their behaviour can be verified to match human-specified requirements and that people are involved in overseeing and checking system decisions.
RAS systems as marketing tools
The operation of these technologies within the home, for instance, in seeking to sell additional services and products to users should be subject to regulation as already applies to existing forms of online and distance selling. There may be a need for additional forms of control given the growing pervasiveness of AI technologies in people’s daily lives.
Respect, dignity, and the right to human care
As people develop increased needs for care there is a risk that they can be treated simply as recipients of support and less as the individual people that they always are. Human subjectivity and experience is changed by some conditions such as dementia but not necessarily diminished. By helping with some of the more intimate and personal aspects of daily life, RAS technologies can assist people to maintain their dignity and autonomy, nevertheless, we consider that it is essential that there is always a human contribution in social care. Indeed, we advocate legislation to protect the right to direct contact with human carers alongside the development of robotic and AI technologies that can support people to live more independent lives.
Impact on carers
Whilst RAS may impact on how social care is delivered in the future, we do not believe that it will replace human jobs in the care sector. First, as technologists who are trying to understand the challenge of care we are very aware of the level of human skill involved in everyday care activities such as helping someone to dress or bathe. RAS can be developed to assist with these activities but they will not match or replace the ability of human carers in the near future. Second, the interpersonal aspects of care such as empathy and understanding are uniquely human. AI personal assistants and social robots may be able to provide a form of synthetic companionship that people may find engaging, but this will never replace human companionship. The UK is faced with a shortage of carers, and care professions are recognised as being poorly paid. The development of RAS for social care should prioritise applications that will relieve the burden on care workers of dull, repetitive and strenuous work, creating a more professional role with a focus on the human-to-human aspects of care. It will be necessary to reassess training needs for some care roles that will in the future require technical knowledge related to customising and deploying RAS technologies. We are likely to see the emergence of a new group of care professionals whose skills include knowledge and understanding of controlling RAS technologies with varying levels of autonomy.
Nationally and internationally a number of efforts are underway to develop governance frameworks for AI and robotics. Within the UK, an initiative by the Royal Society and the British Academy is considering governance of data , and the Parliamentary STC has called for a Commission on the ethical and societal impacts of AI . With regard to RAS there is the potential to build on the EPSRC principles  and on the European Parliament’s recent effort to develop civil law rules for robotics . Governance and regulation will need to be international if it is to be effective  and not simply promote competitive advantages for less regulated countries; useful leadership in this area is being provided by the IEEE , the World Economic Forum , and the Foundation for Responsible Robotics . We consider that efforts to improve governance of AI and RAS, both in the UK and across the world, are welcome and timely and that the UK can provide global leadership in meeting this challenge.
 Royal Society and British Academy. Data Governance. Accessed 1st July 2017.
 House of Commons Science and Technology Committee. Robotics and Artificial Intelligence. September 2016.
(3). Boden et al. Principles of robotics: regulating robots in the real world. Connection Science, 29(2), 124-129. 2011/2017.
 European Parliament. Civil Law Rules on Robotics. Accessed 1st July 2017.
 Merchant, G. E. and Wallach, W. Coordinating technology governance. Issues in Science and Technology, 31(4). Summer 2015.
 The IEEE. Ethically Aligned Design: A Vision for Prioritizing Human Wellbeing with Artificial Intelligence and Autonomous Systems. Accessed 1st June 2017.
 World Economic Forum. Global Risks Report 2017. Part 3: Emerging Technologies. Accessed 1st July 2017.
 The Foundation for Responsible Robotics. Accessed 1st July 2017.