AI: friend or foe to public health?

Artificial intelligence has developed so rapidly over the last few years that almost every sector has had to consider how it might affect its work. Public health is no different.

This technology has the potential to make a real impact on people’s health and wellbeing. As it continues to develop, it is crucial that we think about how we use it as a tool for change.

In a recent episode of Spread the Health, Christina Pagel, Professor of Operational Research at University College London, and Martin McKee, Professor of European Public Health, London School of Hygiene and Tropical Medicine, discussed the risks and rewards of the technology, and whether it has the potential to transform our health.

AI feels like “a second industrial revolution”

Pagel describes the advancement of AI as nothing short of historic.

“To me it feels like a second industrial revolution”, she says, explaining that AI is taking away a lot of routine cognitive labour and therefore has the potential to be transformative, in a similar way to the reduction of routine physical labour nearly 300 years ago.

Using AI in public health

From spotting infectious diseases early to encouraging behaviour change, if harnessed correctly, AI could be a powerful tool in the shift to prevention.

“The power of the new AI tools is that they can personalise any message to you […] and that can be really powerful”, says Pagel.

She also explains that AI is good at using lots of different data sets, but it’s what you do with that information that counts: what policies you make, how you implement them, and whether they’ll work.

How much caution should we show towards AI?

We must also consider the risks of unregulated AI, particularly when driven by commercial incentives.

Pagel emphasises that AI can personalise messages “at pretty much zero cost”. While this can support public health campaigns, it can also spread misinformation and cause fearmongering.

McKee points to the broader risk of choice:
 

It’s important that AI tools are regulated properly to ensure that they are used for positive change.

Does AI further cognitive bias?

There have been plenty of criticisms of AI potentially furthering unconscious bias. Pagel explains how using AI on an area you aren’t an expert in can lead people to “trust it completely” and cause human error.

McKee goes one further, stating that “you should never, ever be using [AI] to make decisions about individuals without a human in the loop.” He says we should instead ensure that people consider whether what they are seeing is reality, or a confirmation of their own biases.

What does this mean for the workforce?

An increasing number of people are living with ill health, and many sectors are facing workforce shortages. AI has the potential to strengthen public health, particularly in an under-resourced system. McKee describes the technology as a real “multiplier” for existing public health skills, allowing the workforce to do things quickly and effectively.

However, any new technology is only as good as it’s operator.

We know that the public health workforce welcomes the idea of new technology, but feel they need proper support to be able to use it properly. With the right training, workforces can gain the skills they need to be able to use AI applications safely and effectively. The same applies to early career public health professionals.

Supporting early career entrants

Early career professionals must also be supported to ensure they get the experience they need to enter the workforce.

Pagel notes that in fields like coding, companies are already deciding between investing in junior employees or AI. However, she says, this choice risks companies having no expertise in ten years’ time. The same applies to early career public health professionals.

What’s next for AI in public health?

So, what’s next?

Pagel hopes for a rethink of education and training around AI:

She points out that it’s important that young people are supported to ensure that their skills continue to stay relevant alongside advancements in technology.

Meanwhile, McKee argues that there are limits to what AI can achieve, pointing out that large language models (LLMs) are trained on existing content:

However, both agree that the future of public health and AI will depend on how well we integrate it with human expertise.

This article is adapted from the latest episode of our podcast, Spread the Health. Listen to the full episode below, or wherever you get your podcasts.