How Gen AI can triple your client servicing capacity
Account managers at South Africa’s leading commercial insurance brokerages could be servicing two to three times their current client base if generative artificial intelligence (Gen AI) fully takes hold. This was among the startling realities shared during the latest #InsureTalk event, broadcast online.
Nothing is constant but change
Soul Abraham, Chief Executive: Retail Insurance at Old Mutual Insure, took to the virtual podium to contemplate how technology is changing insurance, with a narrow focus on intermediated commercial and personal lines business. The presenter, who boasts 19-years’ experience in the intermediated insurance environment, is working for a financial services brand that has been plying its trade locally for around 180-years. You would be correct to say that this insurer, alongside its centenarian peers, has seen it all.
The presentation title was somewhat of a tongue twister, promising an explainer on ‘Transforming and enhancing intermediated non-life insurance using Gen AI’. “It is difficult to embrace this type of change because it is fast-paced, invasive, and pervasive,” Abraham said, before describing humanity’s centuries-long dance with development. It all begins with the Industrial Revolution, spanning from 1760 to 1840, during which we transitioned from manual labour to machine-based manufacturing. This era brought the mechanisation of the textile industry and steam engine, among other step-changes.
From 1870 to 1914 we entered the Second Industrial Revolution driven by the development of the internal combustion engine, breakthroughs in steel production, and the widespread adoption of electricity. Communication technologies like the telegram and telephone emerged during this period too. The Third Industrial Revolution, dubbed the Digital Revolution, played out from 1970 to the cusp of the 21st Century. It brought computers, electronics, global telecommunications, nuclear energy, and the Internet. Finally, we enter 4IR, the Fourth Industrial Revolution, dominated by AI, automation, big data, cloud computing, robotics, and many more.
Re-imagining insurance distribution
The presenter argued that 4IR forever changed the distribution footprint of non-life insurers. “We saw the prominence of direct insurers rising across our industry, not just in South Africa, but globally, from around the mid-1990s,” Abraham said. Using his informed best estimates of the distribution landscape, he noted that the direct share of personal lines premium had risen from around 20% in 2007 to 33% in 2014 to 47% today. “Our customers are slowly shifting to the direct space,” he said, before warning that the common retort about personal lines being a commoditised business would not insulate the commercial side.
Direct insurers are chipping into commercial premium despite the complexity of underwriting in this space. Back in 2015, they placed around 4% of the commercial market; today it is closer to 12%. This assessment excludes niche sectors like Agri-crop insurance and large corporate accounts. “Throughout the end of digital era and the start of the 4IR [we have seen] a significant shift in personal lines with a slower but material shift in commercial toward direct,” Abraham said. He noted that Gen AI, also called large language models (LLMs) presented both an opportunity and a risk for the intermediated or traditional insurance industry.
The presentation spun through some excellent plain English narratives to help the audience understand critical concepts. Gen AI was described as the digital version of a human that goes to nursery school, then spends 12-years in school, then five or more years at University, and finally 25-years gathering on-the-job experience. This education roadmap is being achieved through large language models, but at almost unimaginable scale and speed. The first ‘L’ in LLM is key. Abraham said ‘large’ is used in two context. First, the model is trained on a large amount of data, and second, there are a large number of parameters used during training.
Knowledge on tap, instantaneously
So, while a human can read a book and kind of remember the key facts, an AI can read every book, and remember every word, every letter, forever. “The LLMs ‘brain’ capacity and processing capacity is infinitely larger than ours; you can think of each LLM as a super-powered auto-complete,” Abraham said. It is thanks to the sheer volume of data that an LLM is trained on that it can interact with humans at a near-human level, and far more intuitively than the average chatbot deployed online today.
The presenter warned against complacency, using a personal interaction with Google’s Gemini to illustrate how the complex field of commercial insurance is child’s play in the Gen AI world. He asked the LLM for a detailed understanding of how a South Africa based commercial insurer would go about underwriting a restaurant risk. The machine begins its internal process by setting out what it has been asked, then stating how it will go about answering this, then conducting the necessary research, and finally collating that info to produce an answer. Abraham asked two senior underwriting colleagues to review the report, and they did little more than fix a grammar error or two.
Abraham shared this anecdote to dismiss the myth that commercial underwriting is safe from the ravages of Gen AI and / or direct insurers. He painted a future in which the likes of Old Mutual Insure would purchase their own LLM license, and then set about integrating group IP such as rating and underwriting philosophies into the extensive LLM capabilities. The result will be an extremely powerful underwriting-focused Gen AI. “If the intermediated industry fails to embrace digital transformation, then we will be accelerating the trend away from intermediated towards direct, in both the personal lines and commercial spaces,” Abraham said.
The biggest risk centres around relevance
The danger is not that Gen AI will replace intermediated insurance, but that the sector will lose relevance by failing to match the pace and precision of direct insurers already deploying it. Brokers and insurers risk falling behind on pricing, risk modelling, and customer experience to name a few. Clients are already interacting with Gen AI across other service platforms; they will come to expect the same ease, responsiveness, and personalisation from their insurance providers. The way forward is through decisive partnerships.
Brokers and insurers must jointly harness Gen AI to automate administrative tasks, enhance fraud detection, personalise offerings, and strengthen underwriting, among other tasks. The true opportunity lies in improving broker-led experiences and creating operational efficiencies that benefit everyone.
Gen AI is changing the world
As Abraham concluded: “Gen AI is changing the world; insurers and brokers will have to partner to harness its power to grow the intermediated sector.” The reward for keeping pace could be profound, with the potential for brokers to double or even triple their client base by leveraging the power of Gen AI and other emerging technologies.
Writer’s thoughts:
AI, Gen AI, and other emerging technologies are pervasive and here to stay. Do you think brokers and traditional insurers are adopting this tech fast enough to close the gap between themselves and the direct guys? Please comment below, interact with us on X at @fanews_online or email us your thoughts [email protected].
