Artificial intelligence – a disruptive technology for the (re)insurance industry
AI is currently experiencing incredible hype. Applications such as generative AI, which produce text, music or images, have become widely accessible in a very short time and are now used across all areas of daily life.
Loading ...
Artificial intelligence is permeating both everyday life and the business world – and the (re)insurance industry will be impacted across the board.
The AI environment is changing rapidly
Artificial intelligence (AI) aims at building intelligent entities which imitate human intelligence. The main areas of AI are knowledge representation and reasoning, perception, linguistic ability and machine learning. AI is increasingly being adopted across many industries and will have a huge impact on business strategies and success, new products and customer expectations.
Fields of application in the insurance industry
Existing examples of insurance applications are chatbots or product recommendation systems. A chatbot is a computer program that allows users to conduct a conversation via speech or text. The computer program replies to the user within milliseconds because the rules of the communication are predefined. In contrast to call centres, chatbots can handle large amounts of user requests and can respond in real time. One possible application scenario is a FAQ chatbot that answers frequently asked questions from customers so they are not required to chat with a call centre agent. A product recommendation system seeks to predict and show the items a user would like to purchase. AI algorithms allow insurers to analyse individual clients, their habits, likes and dislikes, claims history and of course their financial behaviour. They can draw conclusions quickly about how to engage with them and which products they are likely to purchase.
Challenges for the (re)insurance industry resulting from AI
For artificial intelligence to change business, it needs to be fuelled with quality data. Even machine learning and deep learning technology – which can make decisions and adjust their actions without explicit programming – need exposure to data in the first place. These require data to be both carefully selected and meticulously prepared.
Bias in data – a hidden risk in AI deployment
Another sometimes even more critical issue is data bias. The problem of bias in machine learning is likely to become more significant as the technology spreads to critical areas like medicine and law and as more people without a deep technical understanding are tasked with deploying it. Some experts warn that algorithmic bias is already pervasive in many industries, and that almost no one is making an effort to identify or correct it.
Regulatory activities
Several initiatives and projects are working on automated underwriting and claims handling systems. The main issues in this field of AI application are data quality, data security and compliance with regulatory implications. The EU General Data Protection Regulation (GDPR) requires meaningful human interaction during otherwise solely automated processes. According to the GDPR rules, insurance applicants have the right to receive a decision not solely based on an automated decision.
Augmentation instead of replacement
Artificial intelligence is commonly sold as a method which automates whole processes. After an evaluation phase, people may be disappointed that AI cannot automate an entire process completely error-free. Hence, one should set the right expectation - AI will not replace all human labour in all processes. In most cases, it automates or optimises a part of a process to free resources for subtasks which require more human intelligence. Finding the right balance between automated and human processes is a major task while implementing AI processes.
Underwriters have to deal with a changing risk landscape in relation to AI applications. Understanding the combined effects of several technologies and their contribution to the overall risk is a major challenge for (re)insurance underwriters. The evolving environment will lead to exposures that are more complex.
Opportunities vs. caution in AI use
AI already presents excellent opportunities, but also significant challenges. Applications like ChatGPT have seen remarkable user growth in a very short time. Despite the current euphoria and optimism, it is essential to recognise the limitations and potential risks of AI. Going forward, striking a balance between critical use and blind trust in AI outcomes will be crucial.