Building a framework for the ethical use of AI in adult social care

In February 2024, a group of 30 organisations – academics, technologists, carers, people who use care and support, and advisory bodies – came together at the Institute for Ethics in AI at Oxford University to discuss what the responsible use of generative artificial intelligence (AI) would look like in adult social care.

Together, we released the Oxford Statement, which stated that “we urgently need to develop a shared, co-produced framework to underpin the ‘responsible use of generative AI in adult social care”. The following article explains our work to date.

In late 2022, OpenAI released ChatGPT and suddenly it seemed like the whole world was talking about generative AI. It soon became clear that the world would never be the same again.

We knew that this was, and continues to be, a developing story that affects us all, including everyone in the social care community. This includes:

• People requiring care and support
• People working in care
• Care organisations
• Care representative organisations
• Tech and software providers to the care community
• Statutory and regulatory bodies

We wanted to develop a learning and action network for the responsible use of generative AI in the care sector, and we wanted to ensure that past mistakes were not repeated, mistakes where new technology and innovations were used in care without involving all the relevant groups involved in the care community. We wanted to learn from the past.

As a team, we knew that we could convene a wide range of people from different groups in the social care sector, and so we set ourselves the task of getting representatives from different stakeholders to a round table in early February 2024. We found a willing host in Reuben College, Oxford and the associated Institute for Ethics in AI.

As the three joint co-ordinators, we led a day’s working group of about 30 people to bring together ideas, and from that group we were able to produce the Oxford Statement on the responsible use of generative AI in Adult Social Care as a starting point. The focus on the word responsible was a key part of discussions, and it was asserted that when talking about AI in a health or social care context, it should be a given that AI should be used in a responsible way.  

Following the roundtable and our commitment to a six-month timeline of work, we have continued to move at pace. This has included setting up a steering committee to oversee our work, launching our public facing webpages, and facilitating a growing number of working groups. Our joint goal is to collaborate on guidelines and suggested uses of generative AI in social care. We want to start a movement encompassing the responsible use of AI focussed on achieving the best outcomes for people who require care and support.

To achieve this, it’s vital that our work is collaborative, open, and developed from the ground up. This is why we have the following working groups working towards our goal:

• Care Workers: The Care Workers Charity has co-facilitated a day’s workshop for care workers who have drafted their own Care Workers’ statement on the responsible use of generative AI
• Care Providers: The care provider network is administered by Digital Care Hub to explore what guidance and resources organisations need
• Tech and software experts: The expert reference group for technology and software companies is meeting to work on eh principles of using AI in social care
• Ethics and Principles: The Principles working group, led by Donald MacAskill of Scottish Care, has been developing a framework of principles to support care organisations adopting AI
• People with Lived and Living Experience: Several co-production groups of people with lived and living experience have also been set up and are meeting both virtually and in-person

We stress that this is a movement rather than a piece of work. It is evolving, and we are aiming to bring the work of all the working groups together to present to the care sector a way of working with AI which can enhance people’s lives. We are setting in motion a way of working, which we think the whole care community can embrace, enhance and develop.

The Oxford Project: The Responsible Use of Generative AI in Social Care is a work in progress. We want to include use cases of how people in care are using generative AI responsibly to achieve better outcomes, so come on the journey with us.

Please contact us with your thoughts and experiences and find a way to get involved: Contact Us on the Digital Care Hub or email help@digitalcarehub.co.uk.

Or you can contact the authors of this blog via the links below.

Katie Thorn, Project Lead at the Digital Care Hub: Email, LinkedIn.

Caroline Green, Institute for Ethics in AI, University of Oxford: Email, LinkedIn.

Daniel Casson, Managing Director at Casson Consulting: Email, LinkedIn.