AI: An unknown risk for life insurers

A lot has been written about artificial intelligence, and its ability to transform industries by improving analysis and improving process speed for many tasks historically performed by people.

Whilst there are clear advantages, both commercial and otherwise, there is a potential cost in terms of a negative impact on people’s skills. In this article we explore some recent research on this very topic, and why it matters to insurers.

The research

Recent research by MIT focused on finding out the cognitive cost to humans of using artificial intelligence, specifically Large Language Models (or LLM). The study found that LLM use:

• is likely to decrease learning skills;
• may diminish critical thinking capabilities and lead to decreased engagement in deep analytical processes;
• may inadvertently discourage lateral thinking and independent judgement;
• reduces cognitive load by 32% compared to software-only users.
• leads to less deep engagement with the material;
• reduces cognitive load, which in turn leads to a shift from active critical reasoning to passive oversight; and
• when used in essay writing leads to homogenous results.

The research says:

“This cognitive offloading phenomenon [from LLM use] raises concerns about the long-term implications for human intellectual development and autonomy”.

However, the research also found that:

“LLM users are 60% more productive overall and due to the decrease in extraneous cognitive load, users are more willing to engage with the task for longer periods, extending the amount of time used to complete tasks.”

These two findings perfectly highlight both the advantages and disadvantages of increased LLM use, although it is worth noting that the research has not yet been peer reviewed, and its sample size is relatively small. 

Loneliness and ChatGPT use

In separate studies, from OpenAI and MIT, links have been found between loneliness and ChatGPT use. Generally, the more time users spent ‘talking’ to ChatGPT, the lonelier they felt. Again, these studies have not yet been peer reviewed, although they provide at least anecdotal evidence of perhaps another negative impact on human mental wellbeing and cognitive function caused by AI use. 

Is there be a link between LLM and dementia or other health conditions?

Loneliness a risk factor for dementia; feeling lonely increase the risk by 31%. This increases the risk at a magnitude similar to being physically inactive or smoking).

Studies also suggest that keeping the brain active could help delay the onset of dementia. How this could intersect with the long-term impact of increased use of LLMs is not known.

What does this mean for insurers?

As LLM use has only recently started with the general population, the long-term health impacts on humans are not known. However, the impact of artificial intelligence on human health will be of interest to insurers, particularly life insurers. The cognitive consequences of LLM is an area where further research is needed.

In the UK, health insurers do not typically cover the long-term costs associate with dementia. Although some diagnostic tests and consultations may be covered.