Artificial Intelligence (AI) technology has the potential to be a transformative force in insurance underwriting, with the most immediate and obvious benefit being the ease and efficiency it can bring to the process.
“Imagine a system that inputs key underwriting parameters, like risk appetite and insurance risk profile, to automatically calculate a rate aligned with your reinsurance treaties and excess-of-loss structures. This level of precision and automation is game-changing,” says Christopher Appanah, General Legal Counsel and Partner Principal at iTOO Special Risks.
He explains AI can synthesise vast amounts of underwriting data and present actionable options, freeing up underwriters to apply their expertise, judgment and experience, rather than getting bogged down in administrative tasks.
“This shift is not only about productivity but also about minimising human error. Errors in underwriting files can lead to unfair claim rejections or improper payouts, affecting both the insured and the insurer,” he adds.
Additionally, AI can unlock the power of data, as from a data analytics perspective, insurers often sit on heaps of historical data, some dating back centuries. That data holds immense value in crafting sophisticated risk assessment tools.
“With AI, we can drill down to granular levels – region, city, even street name – to identify top risks and underwriting trends. We can analyse region-specific underwriting behaviours, claims histories and gross written premium (GWP) performance to refine our approach,” says Appanah.
“AI can ingest all of that, apply your rules and deliver insights that are not just fast, but deeply valuable. It is not just about automation; it is about intelligence.”
Underwriters have often been seen as desk-bound analysts, but with AI taking care of administrative tasks, they can focus on building relationships and engaging with clients and brokers. In a regulated environment with increasing governance demands, AI not only boosts efficiency but also enhances human connection.
Appanah highlights that insurance is fundamentally a relationship business, where many conflicts are resolved through trust and dialogue rather than policy wording. AI aids in maintaining the human touch by freeing up time to strengthen these relationships.
However, South Africa’s insurance industry has not fully tapped into AI’s potential, especially in underwriting and claims. While some players use AI to generate rates based on minimal input, the lack of human connection is a major flaw. Loyal, low-risk customers can be penalised by rigid algorithms, and when issues arise, the absence of human support erodes trust.
“Others are using AI more strategically for fraud detection. These systems operate like spider networks, connecting to industry databases, such as the South African Insurance Crime Bureau (SAICB), to flag individuals with a history of fraudulent claims. But beyond that, I have not seen any breakthrough applications of AI in South Africa’s insurance sector. And that is a huge gap,” says Appanah.
The hesitation often stems from fear about job loss, high costs and uncertainty. But AI is not here to replace people; it is here to enable them. It can streamline operations, enhance customer service and free up time for relationship-building, which remains the bedrock of our industry.”
He says that while many companies talk about “incorporating AI”, few truly understand its full potential. Most are still using it for basic tasks without realising how deeply it can transform their business.
“This is not just a technical skill; it is a strategic capability. And while South Africa’s insurance sector has not fully embraced AI yet, the momentum is building. We need to ramp up, shift the narrative from fear to enablement, and start using AI not just to automate, but to elevate,” Appanah concludes.