Artificial Intelligence has a marketing issue…the problem is that our perception is heavily influenced by science fiction’s portrayal of AI as highly intelligent, autonomous, and possessing near-human level capabilities. But if the reality is that AI is missing fundamental capabilities such as common sense, emotional intelligence, creativity, and the ability to understand and learn from context, where can we find use for it in subsidence claims?
Major advances have been made in machine learning but they’re in niche and narrow fields – you’ll be familiar with breakthrough gaming headlines such as Chess and Go. But there is still a chasm between where AI is today and where we need it to be to solve all subsidence claim challenges.
So, where is AI today and how much does it understand subsidence claims? The best way to find out is to simply ask it! For this exercise, we posed the question ‘Explain the steps in handing an insurance claim for damage to buildings in England caused by subsidence’ to ChatGPT, a text-generating AI bot. Are you ready for the answer?
Here was what ChatGPT had to say:
Step 1: Contact your insurance company as soon as possible to report the damage and start the claims process.
Step 2: Provide your insurance company with documentation, such as photographs, of the damage caused by subsidence.
Step 3: The insurance company will send out a loss adjuster to assess the damage and determine the cause of the subsidence.
Step 4: If the subsidence is found to be covered under your policy, the loss adjuster will estimate the cost of repairs.
Step 5: Once the cost of repairs has been agreed upon, the insurance company will pay out the claim to the policy holder, which will typically be used to pay for the repairs to the building.
Step 6: If necessary, the policy holder may need to engage the services of a building contractor to carry out the repairs.
Step 7: After the repairs have been completed, the policy holder should provide the insurance company with documentation, such as receipts, of the repairs carried out.
Step 8: The insurance company will then check the repairs have been completed to their satisfaction and will release the final payment.
How would you score the answer, out of 10?
At first glance, it’s mighty impressive. ChatGPT clearly understood the question and correctly structured the answer into different steps. The answer is articulated well and it’s fair to say that to any non-subsidence claims expert would be content with the response. However, here in-lies the problem, any claims expert will instantly see the gaps in knowledge and be able to pick holes…
Although ChatGPT used the British term ‘adjuster’, the settlement steps have the flavour of a North American property claim. There’s also no reference to the insurer instructing repairs – common UK practice and the method preferred by FOS when there’s structural damage. But the biggest omission is consideration of mitigation, a major oversight, leading to ChatGPT’s unsuccessful job application for Head of Claims Operations at any insurer.
Five out of ten seems a generous mark although not bad when considering it took just 15 seconds to produce the answer. If you ask it to explain 'mitigation' when handing an insurance claim for damage to buildings in England caused by subsidence, the answer comes in 20 seconds, but is still only a 5/10.
But are we being unfair here by taking a fish out of water? This is where we circle back to AI finding niches to deliver extreme efficiencies and where, in the right environment, AI thrives. We’re already seeing it provide deep insights into both structured and unstructured data analysis to help insurers make better and more informed decisions. And that’s just it, AI is another tool in the ever-expanding toolbox insurers can delve into to differentiate themselves and build competitive advantages in the digital landscape.
Like any tool, it needs to sit between creativity and expertise to deliver the best results, a bit like a camera…we all have cameras on our phones but we can’t all take award winning pictures. You first need the idea of what to capture and then you need the expertise to know how to capture it. And AI tools that are available today are no different.
The good news is that no one understands the power of using the right digital tools for the job more than us. And it’s our experience and expertise combined with our digital technology that continues to set the industry benchmark for subsidence claims management.
So, whilst AI may not be ready to take over subsidence claims, it’s certainly ready to take part.
360Globalnet, the market leader in digital subsidence claims, marries together digital and human touch to deliver the most efficient subsidence claims journey on the market. Get in touch to find out more.