ARTIFICIAL INTELLIGENCE (AI)
Updated: Jun 23
by Navdeep Arora
The New Magic in Insurance
Artificial Intelligence (AI) is revolutionising different industry sectors through unprecedented improvements in accuracy, speed and the cost of doing business. AI has become a buzz word in Insurance. Improvements have been broad and deep: from tailoring products to customers to early fraud detection and automation in underwriting and claims management for core insurance products (e.g. term life, personal accident, auto and home insurance). With the magic of AI now established, insurance will continue to adopt and refine its AI-related capabilities to transform how insurance meets our evolving needs – and impacts our lives.
AI technology has progressed immensely in the last few years, and insurers are embracing its ‘new and improved’ facets. Such advances range from natural language processing, computer vision technology, machine learning to neural networks. Specifically, robotic process automation has enabled insurers to use AI to empower agents, brokers and employees enhance the customer experience via personalization, such as offering services like individual risk-based underwriting processes and faster claim settlement. These advances are multi-beneficial – and magical.
At the core, the magic of AI will revolutionise four core capabilities that underpin and underlie the insurance value chain. These are:
Symmetry of Information: AI will help reduce the asymmetry of information that exists between the insurers. Historically, insurers sought to extract information via questionnaires, observations, demographic statistics and claims losses to conduct actuarial analysis and rationalise future underwriting and pricing decisions. Currently, Insurers use AI to contextualise needs and risks based on the most accurate and real-time data gathered from multiple external sources (now available) in combination with historical data, and losses. And this endeavour is set to continue to evolve as insurers look to futurise: using data to predict behavioural patterns. According to Cisco, 500 billion sensory devices with 4 -5 signals each will be connected to the IoT by 2030, implying about 250 sensory data points per person on average. This will enable insurers to offer on-demand, pay-as-you-use (PAYU) policies tailored to individual needs and risk profiles at much lower costs. Insurance start-up Laka (United Kingdom) is using AI to tailor insurance policies and lower costs to bikers and e-bikers by 20-30% and at a time when the mobility trend is spiking due to social distancing in the current and post-Covid-19 economy.
Better Risk Mitigation and Prevention: AI is enabling insurers to better mitigate and prevent risks and adverse events alongside traditional protection. In the health insurance sector, insurers like Aditya Birla Health Insurance (India) are using wearables, such as fitness trackers and heart rate monitors, to collect data and monitor health needs to provide customers with the best health insurance coverage options. Insurtech start-ups like MariaHealth (The Philippines), and Collective Health (United States) are compelling examples of companies that use machine learning to maintain AI-powered profiles of its members to identify risks and provide appropriate healthcare resources to them. In car insurance, innovative start-ups like Gypsee (India) are looking beyond relying on basic driver and vehicle information to craft insurance policies by using telematics to get real-time driving data to predict behaviours, prevent accidents and incentivise safe drivers. Likewise, insurers are working with claims technology providers like MotionsCloud (Germany) to use AI models based on computer vision – the science that enables machines to extract meaning and context from visual data – to assess vehicle and property damage and associated repair costs. MotionsCloud also enables policyholders to report the event photographing their damaged car or property. This information is then submitting to an AI-automated damage estimator that assesses the damage and ensures a quick and seamless claim settlement that is hassle-free for both the insurer and the insured. Many insurers already use machine learning to detect and prevent fraud, and despite the complexities of big data and the overabundance of unstructured data in particular, the unfathomable amount of data available today can be used to train machine learning algorithms to analyse patterns and separate legitimate claims from the fraudulent ones.
On-demand Customer Service: AI will support customer service that is on-demand and always available across all channels (online, mobile and telephone) to improve customer access, acquisition and retention. In particular, advances in Natural Language Processing (NLP) have empowered insurers to extract information from unstructured data (e.g. textual data from documents, chat logs, emails and social media). Insurtech start-ups like Democrance (UAE) are using NLP to help insurers pro-actively service the needs of their policyholders. Similarly, Digital Fineprint (United Kingdom) is improving customer acquisition and retention through AI-based social media data analysis. To further enhance customer experience, customer service chatbots have been being introduced by several insurers like AXA (France), Lemonade (United States) and MetLife (United States). Chatbots are a digital service available 24/7 that provide advice ranging from basic support for common enquiries of policyholders (e.g. policy terms, support and advice) as well as online pricing tools for prospective clients (e.g. product comparison, basic insurance package creation).
More Efficient Operations: lastly, AI is enabling insurers to improve the efficiency of their operations through automation across core activities. These include processes such as data management, customer service, underwriting, claims and reinsurance placement as well as in administrative processes. Though somewhat in its early stages, as more insurers adopt AI for many reasons listed above, and more, greater accumulation of data will help them continue to lower costs, increase competition and, ultimately, produce and deliver better products and prices for consumers.
Ahead: New Challenges and Opportunities
AI brings both new challenges and new opportunities for the insurance sector. Internally, there is a desire and need for insurers to look inwards and educate, empower, upskill and train their insurance professionals to work with data and AI-related capabilities, while also addressing cultural, behavioural and mind-sets related obstacles that might hold them back. Conversely, insurers also need to look outwards to proactively identify new and emerging risks, data and data analytics and the many opportunities arising from short-term events like the current COVID-19 pandemic as well as long-tail impacts such as the coronavirus’ impact on global supply chains, rising mobility trends and the growing GIG economy. The new magic of AI for insurance is in its early stages, and we have yet to realise the full potential of its magic. But there is strong consensus, and noticeable effort, that the opportunities AI promises to unveil and unlock are compelling, unprecedented and beneficial for all.