Artificial Intelligence is applied when a machine mimics cognitive functions that humans associate with other humans, such as learning and problem solving. The rules-based system components of AI are a "natural fit" for many transport businesses, particularly marine insurance underwriting given the structured nature of the questions and answers required. Below we look at AI and its influence on Marine Insurance. One important use of AI in insurance is to automate administrative tasks through "autonomics". The software is used in high volume, rules-based work and can replicate decision-making processes and apply them quicker and more efficiently than human operators.
The insurance industry generally is notorious for its outdated processes and fails to leverage new technologies that are available to it. Insurers would benefit from automating routine tasks but many are yet to even begin with reviewing their processes, simply because their legacy systems are complex transactional platforms that are unable to integrate with and make use of new technologies. There is evidence that marine insurers are beginning to embrace AI and make the necessary changes to their organisations to enable the use of AI-related technologies. These help fleet operators and marine insurers to improve the reliability of vessels and lower their costs of operation, whilst also facilitating safe operations and ultimately lowering risk.
It is anticipated that AI will have a huge impact on all aspects of the marine insurance industry. Advanced technologies and data are already affecting distribution and underwriting, with policies being priced, purchased, and bound in near real time.
The experience of purchasing insurance will be faster with less active involvement by the insurer and the customer. By 2030, enough information will be known about individual behaviour for completing the purchase of a Hull & Machinery, cargo, or general liability policy will be reduced to minutes or seconds. Smart contracts enabled by blockchain will instantaneously authorise payment from a customer's account.
By 2030, claims processing is likely to remain a primary function of carriers, but head count associated with claims will be reduced by 70-90% compared with 2018 levels. Advanced algorithms will handle initial claims routing, increasing efficiency and accuracy.
Underwriting and pricing – Manual underwriting is predicted to cease to exist for most personal and small-business products across casualty insurance by 2030. The majority of underwriting is likely to be automated and supported by learning models powered by internal data as well as a broad set of external data accessed through application programming interfaces and outside providers. Information collected from devices provided by mainline carriers, reinsurers, product manufacturers, and product distributors will be aggregated in a variety of data repositories and data streams.
These information sources will enable insurers to make decisions regarding underwriting and pricing, and will enable proactive outreach with a bindable quote for a product bundle tailored to the buyer’s risk profile and coverage needs.
Therefore, whilst there will be employment cuts, AI-related developments will create a new area of the industry that will employ significant numbers of people in analytics, reporting, marketing and software development. Insurers must train for the future and not hold onto the training programs of the past. A key concern in introducing new technologies will be in convincing the customer that automation is not a Trojan horse for denying claims. However, there are arguments that AI could actually make the insurance industry more trustworthy by using technology and behavioural science to create a faster and more transparent service.
The insurance industry is ripe for automation intervention as it revolves primarily around analysis and processing of information. Any shift towards automation should deliver significant cost savings given that a significant portion of an insurer's cost structure is devoted to human resources. Additionally, insurers could cut their claims processing times down from months to minutes, or even seconds. Machine learning is often more accurate than humans; insurers could therefore also cut down the number of denials that result in appeals they may ultimately need to pay out. AI can be used to quickly understand the potential impacts of events around the world as it is able to take unstructured data from sources such as social medial, weather, real-time securities fees, and emails, to analyse them in real-time to predict scenarios and evaluate impacts. For example, claims resulting from severe weather events. AI will also assist insurers in the costly ongoing battle against fraud: it can sift through vast amounts of external and internal data to detect anomalies and fraud-related patterns such as duplicate claims, learning each time an event leads to a confirmed or false case of fraud.
Most marine underwriters are "keenly" aware of the need for better data analytics but struggle to find data of sufficient resolution and accuracy to use when selecting risks and setting prices. Until recently, there has been a lack of tools to enable an underwriter to select and monitor vessels and build a high performing, well-diversified insurance portfolio. AI may represent a solution and is therefore likely to change the industry drastically – if it is embraced. There is evidence of a strong interest among marine insurers to engage AI, paired with an interest of insurtech investors who have identified marine as the focal point of the technology revolution affecting insurance. This focus is due to the high value of the commercial marine insurance market, the emergence of a wealth of data within the industry (albeit fragmented at present), and the fact that it sits at the heart of global supply chain logistics. Commercial marine insurers have never had so much data available to them, but many do not have the tools or skills set to take advantage of it.