Dean Collins, human-centred AI expert at PA Consulting, says we need to re-think the ‘people aspect’ if AI’s benefits are to be felt by care providers and communities.

The Government’s AI Opportunities Action Plan hopes to position the UK as a global leader on the AI playing field. But what does it mean for local government leaders on the ground? AI is key to the Government’s aim of delivering more efficient services through local government. We know that local councils are exploring how AI can help with over-stretched services, particularly in social care – but we need to re-think the ‘people aspect’ if AI’s benefits are to be felt by care providers and communities.

Time spent or time saved?

AI promises the opportunity to save time from more repetitive, mundane tasks. There’s a widespread assumption that technical training will help teams and departments unlock the value of AI. However, the reality is more nuanced – time-saving alone is not a benefit; it’s how people spend the time that drives the benefit.

For example, a case worker who would previously write up meeting notes can instead use recording software to draft notes in real-time and quickly review them. But there’s little point in achieving time-savings if those hours aren’t then put to good use, such as on more citizen-facing activity. Social care leaders should record how many hours AI-enabled actions like this save – and then set out how exactly staff can use that time. As employees are released from more administrative tasks, there will also be a pressing need to re-train people for the shift in roles.

Put simply, for AI to successfully enhance the way local services are delivered, there needs to be a wholesale review of the future of work and training people to fill future roles. Taking a whole-human approach that prioritises outcomes and soft skills over process-simplification will be central to unlocking efficiency and better value for money.

Benefits of AI

The use of technology in social care may not be new: predictive texting, autocorrect in Outlook, and email drafting in Copilot have all reduced administrative burdens on social workers. But AI can take this further by helping identify the best option for an individual, condense vast quantities of information, and deliver personalised services.

For example, why not use this information to make predictions on the types of care that a service user will likely need in future, providing social workers with options to build a care package? AI can help coordinate service deliveries across large numbers of people, providing the services they need before they may even need them, and turbo-charging the services’ impact. This can be complemented by drawing up guidelines for the types of service users who will particularly benefit from more face-to-face, intensive personal support.

Other benefits of AI can be seen in assistive care technologies, which can offer a ‘whole view’ of the service user and improve the service they receive. We’re seeing smart sensors that assess a person’s ability to live independently post-hospital discharge, enabling social workers to make informed decisions about care packages. In another case dealt with by Argenti, PA’s social care technology service, a twenty-year-old client was living with a degenerative condition and received care in bed. Their parent struggled to continue as a carer and felt obliged to seek nursing care at great cost to the local authority. Argenti carried out an in-person assessment and installed AI-based voice assistant solutions, instantly reducing the need for a carer and improving quality of life. All this for a fraction of a cost of a nursing care home- and custom-made environmental controls.

Navigating AI overreliance

There are, however, risks of overreliance on AI. AI is a decision-making tool, not the decision-maker. This means AI should not make choices, but instead provide brain power for humans to make a better-informed choice.

Its important social workers can use the information that AI has captured to identify and help people most in need. They need to be able to draw on their experience to build a rapport and ensure that the support provided has the best possible impact. To do this, social workers should be trained to challenge AI’s suggestions and not just follow a process which could deliver an impersonal, potentially costlier service that is not the right fit.

This ability to provide a tailored provision is a skill that should not be lost. The increased use of AI requires safeguards and clarity, so care workers can understand and fulfil their roles in the new environment. AI-driven changes are about more than just technical developments. We cannot pre-define every possible action a care worker will need to take – nor should we. We should be equipping people with the knowledge, skills, confidence and capability to make the right decision, at the right time, in the best interests of the individual.

Achieving this will require a mix of technical skills, ability to use the technology, social care specific capabilities and soft skills such as communication and leadership. For councils, this means wrapping AI around specific end-to-end services or use cases to focus on specific outcomes that need delivering – such as spending more time with service users by reducing time spent note-taking. In this respect, human-centred AI adoption offers huge opportunities to deliver services that improve people’s lives.