AI in Health and Social Care Forum Reflections

Global context

Artificial Intelligence (AI) is a central part of national strategies and global economic competition, and it is not a new concept. AI has existed in theory since the 1950s, but recent advancements in large language models and tools like ChatGPT have made it far more visible and widely used. These developments have created confusion about what AI is and what it isn’t. The roundtable focused on practical uses of AI that analyse large datasets to provide insights to help humans make decisions, rather than hypothetical Artificial General Intelligence, which does not exist yet.

As investment and interest in AI grow, so do the risks that come with it. Many AI-enabled tools and technologies rely on a small number of companies that make advanced computer chips and run large data centres. If those companies faced disruption, increased prices, or restricted access, it could have a significant impact on the end users. In addition, AI systems can consume large amounts of energy, making them costly to run and raising concerns about the impact on the environment. These factors converge to influence where AI can be used, how resilient it is, and who can afford to use it.

There is also a wider societal question: how should we decide where and how AI is used? As the Brookings Institution notes, technological change shouldn’t happen just because it is possible or profitable. Society should shape the use of AI in ways that align with public interest, especially when it affects health and care.

Source: The Global AI Index, Tortoise

Practical applications in health and social care

Participants stressed caution in a saturated market. Many AI products look great but rely on generic off-the-shelf models underneath. AI works best when it is applied to specific and well-defined problems. Its use should also be supported by high-quality data, clear oversight and good governance frameworks.

Examples discussed include:

Helping doctors by highlighting the latest research and evidence to inform decisions.
Detecting patterns in datasets to support diagnosis, treatment, and early intervention.
Supporting prevention and personalised care, such as tailored exercise or lifestyle advice, to reduce pressure on frontline services.
Care-focused technologies that promote independence without replacing human relationships
Supporting Security Operations Centres (SOCs) by detecting suspicious activity, reducing false alarms, and helping teams protect sensitive patient and operational data.

The shared principle across all the examples discussed was that AI should handle tasks faster or more accurately than humans, while humans retain complex judgment, empathy, and personal connection.

Source: Bioengineering (Basel). 2024 Mar 29;11(4):337. doi: 10.3390/bioengineering11040337

Challenges around data, trust, and ethics

AI depends on accurate and structured information to work effectively. In the health and care sectors, data records are held across multiple legacy systems and organisations. Cleaning, linking and managing this data is time-consuming but key to creating more reliable AI outputs.

There is also a risk of over-relying on AI. Using AI tools outside of their intended purpose can lead to errors that can affect anything from clinical decisions to patient independence and wellbeing. AI must be carefully monitored to ensure that it is not learning biased or misleading patterns, attributing unrelated lifestyle factors to a particular disease, for example.

In their everyday lives, people are generating more personal health data through apps, wearables, and in-home devices. This raises questions about who owns the generated data, how it can be shared, and how consent for sharing is managed. Protecting patient privacy while allowing effective use of data is an ongoing challenge.

The Oxford-led Digital Care Hub has published practical guidance on responsible and ethical use of generative AI in social care, as an example:

A deep dive into challenges around generative AI adoption

Source: The ingenuity of generative AI report, IBM, 2024

Skills, talent, and the workforce

AI is changing the face of the modern workplace. Its use can boost productivity, but it may also displace roles, reduce opportunities for junior roles and affect leadership pipelines as a result. The consensus is that AI should support, not replace, human workers. Skills like communication, critical thinking, and effective human-AI collaboration will have a greater emphasis placed on them in the hiring process.

Global competition for AI talent is intense. Countries like India have a large AI talent pool, leading some organisations to outsource work there due to lower cost and availability of specialist skills. To remain resilient, the UK must invest in local talent, training and retention.

Looking ahead

Across health and care, AI can enable more personalised services, improve decision making and extend support, without removing the human element that underpins patient trust.

Key considerations for the future include:

Making AI tools and technologies accessible and inclusive for those most in need.
Improving existing services while lowering costs.
Addressing challenges like broadband, energy, and sustainability.
Breaking down data silos between organisations.
Investing in domestic AI talent to reduce reliance on overseas expertise.

The overarching message was that AI in health and care must be deliberate, ethical, and collaborative.

With thanks to:

Find out more about HCR Law’s Healthcare Specialism