AI Adoption Challenges in Commercial Insurance: Industry Leaders Discuss Solutions
Artificial intelligence (AI) models, powered by advanced technology and large language models (LLMs) like ChatGPT, are revolutionizing the insurance industry. Companies such as Zurich and Artificial Labs have already started experimenting with chatbots and AI for claims processing and customer service. However, implementing AI in commercial insurance comes with its own set of challenges, according to industry leaders.
During a recent roundtable discussion, James Burge, Allianz’s head of counter fraud, highlighted the unique complexities of AI adoption in commercial lines. Unlike personal lines, where AI can handle low complexity queries from consumers, commercial lines require a different approach. Whether it’s detecting criminal activity or providing customer service, commercial insurers need to consider AI in alternative ways.
Burge emphasized that when it comes to fraud detection, commercial insurers are not looking for a needle in a haystack. Instead, they focus on filtering out irrelevant information first before identifying potential fraudulent cases. The majority of customers in commercial lines are genuine, so the approach has to be different. Additionally, customer expectations in terms of service and response time vary based on the size of the business.
Adele Sumner, head of counter fraud and financial crime at RSA, expressed concern that AI models might predominantly target personal lines, leaving commercial insurers behind. To increase AI adoption in commercial lines, insurers must develop a robust data collection plan that covers fraud, threats, and trends. Gareth Evans, head of customer success at Shift Technology, highlighted the need to leverage external data to enrich AI models and enable better decision-making.
Cost was another challenge mentioned by Burge. Despite the significant benefits of AI, implementing it at the early stage of underwriting requires investment. Burge suggested that by using AI to identify fraudulent cases from the outset, insurers can enhance business performance and reduce losses. However, understanding the long-term direction of AI adoption in commercial insurance is crucial.
Evans emphasized the importance of starting with commercial claims to develop a comprehensive fraud detection model using AI. By integrating policy detection, automated functionality, and claims fraud detection, insurers can gain a holistic view of customers and their claims history. This level of insight enables insurers to make more informed decisions and actively combat fraud.
In conclusion, AI adoption in commercial insurance presents unique challenges that industry leaders are actively addressing. By taking alternative approaches to AI implementation, developing robust data collection plans, utilizing external data sources, and investing in the early stages of underwriting, insurers can harness the true potential of AI in commercial lines. As the industry evolves, finding the right balance between customer expectations, fraud detection, and cost-effectiveness will be key to successful AI integration in the commercial insurance sector.
