Technological advancements are helping to automate manual work across the industry, and the Open AI platform ChatGPT is one of the elements which could prove to be incredibly useful.

There was a time in the not-too-distant past where self-driving cars, flying drones, and fridges that tell you to order fresh milk seemed mind-blowing, futuristic ideas.  But with the meteoric rise of technology in our everyday lives, such fancies have quickly become old news.  The latest tech to take the world by storm is ChatGPT which gives us the ability to chat with AI. It’s fresh, novel, and intriguing, and many believe it will leave many professions defunct as the world of work is automated even further.  Through API-powered links and other implementations, this new open API platform has many capabilities and the media industry, for example, is already harnessing its powers to their advantage.  For example, Buzzfeed recently signed a contract with ChatGPT to generate articles for its news site.  It’s time for the insurance industry to ask how ChatGPT can help with transport and logistics insurance.  How can this novel new chatbot’s capabilities help solve claims and advise on cargo insurance?

Putting it to the test: What does an AI know about insurance?

Currently, actual insurance starts with a claim. Anyone can write a policy, but not everyone can solve a complex claim. Despite the importance of claims to the quality of service, it is usually seen as a cost by those involved in insurance. It means understaffed claims departments are pushed to the limits with piles of manual work. Can chat AI assist this critical insurance chain?

Let’s establish the fundamental truths. The core of every claim is successful interpretation of the applicable legal regime. In the EU, 42% of shipments are moved by road haulage. The CMR convention regulates the liability limits for cargo by road transport. When asked whether ChatGPT is familiar with the CMR convention for road haulage, the AI scored full marks, giving a succinct, and comprehensive summary of the convention.
The most daunting task of every claim manager is to reject unreasonable demands. You must be polite, precise, and get right to the point, avoiding fancy words like “conclusive actions.” Can ChatGPT aid in writing such an email? We put it to the test again by telling it we had a claim invoice for a new vehicle damaged during the voyage, and that the claim was sent more than 12 months after the delivery date. We provided the AI with all the relevant details of the claim including the correct CMR convention article, and the vehicle details.  We then asked it to write us an email of no more than 4,000 characters to dismiss the cargo claim.

ChatGPT didn’t do an excellent job here because there are areas which can be expanded upon. However, it produced an OK email, sufficient if you just need to get the job done, and at the very least it has provided a substantial framework for the email into which you can simply plug extra details.

Challenging ChatGPT further: calculating liability limits

Currently, ChatGPT’s strengths lie in you writing back and forth with the AI. We wanted to know whether it could assist with more complex claim calculations. So, we created a scenario for a claim with an invoice value and a cargo weight, for which we wanted to calculate the hauler’s liability limit.  We were blown away, the AI gave a correct calculation and even advised on gross negligence and the necessity of cargo insurance.who will still proudly issue your cargo insurance policy, and will also accept no responsibility if you ever have to make a claim.

According to the provisions of the CMR Convention, the liability of the carrier for the loss of goods is limited to 8.33 Special Drawing Rights (SDRs) per kilogram of gross weight. In the case of your lost cargo with a weight of 500 kg, the maximum liability of the hauler would be 4,165 SDRs, which is equivalent to approximately 5,695 EUR at the current exchange rate. It is important to note that the carrier’s liability is limited as per the provisions of the CMR Convention, and the consignor may recover a higher amount if it can prove that the carrier was at fault or acted negligently. In such cases, the carrier may be held liable for the full value of the goods.
It is also advisable to have adequate insurance coverage for the goods being transported, as this would provide an additional layer of protection in case of loss or damage during transit. In conclusion, the maximum liability of the hauler for your lost cargo is limited to 4,165 SDRs or approximately 5,695 EUR, as per the provisions of the CMR Convention.

We continued to challenge Chat GPT: cargo clauses; the difference between ICC and German DTV clauses; the difference between ICC clauses B and C, and so on and so forth.  Here ChatGPT did make a few mistakes but the more we asked questions, the better and better it got at providing accurate answers. After four or five exchanges, providing tighter, better thought-out questions, the AI provided accurate answers. 

The verdict: Can ChatGPT help the transport and logistics insurance industry?

In conclusion, yes, it seems with careful usage ChatGPT can help the insurance industry. Technological advancements are helping to automate manual work across the industry, and the Open AI platform ChatGPT is one of the elements which could prove to be incredibly useful. It will save time and costs in the claims department so long as a professional enters the correct data and is able to direct the AI to the right article and convention. ChatGPT is good with the facts, turning them into stories, emails, letters, and documents; and it’s also good at providing the user with options.  However, it will not advise (yet!) on the final decision in a claims case, e.g., which cargo liability option to choose, but with careful use and rigorous checking of the answers, it certainly has the potential to help the insurance industry in the future.  Watch this space!

Gediminas Dauksa GrECo

Gediminas Dauksa

Group Practice Leader Transportation & Logistics

T +370 616 08451

Related Insights