GenAI could unlock around $2.1-3.2bn in economic value for African insurers, with opportunities across the value chain, according to an estimate by the global management consultancy McKinsey & Company.
To unlock the value potential, insurers can tackle the opportunities fully in one domain at a time—for example, customer engagement and sales, customer servicing and operations, or claims and fraud, says McKinsey in a report on Africa’s gen AI opportunity.
While the majority of current innovations are in South Africa, many other African insurance markets are growing and experimenting with gen AI, including Ghana, Kenya, Morocco, and Nigeria.
South Africa – sophisticated deployments of gen AI in insurance
The South African insurance market, often a reference point even for developed markets in terms of penetration and innovation, is pursuing wider and at-scale applications of gen AI, says the report.
Emerging innovations range from voice bots and enablement in call centres and claims functions to personalised outbound sales campaigns and at-scale hyper-personalised customer engagement through agents and direct-to-client outreach.
For example, one South African life insurer is combining gen AI with behavioural science to equip financial advisers with personalised advice content to engage clients and drive cross-selling, retention, and overall financial well-being. To date, this is one of the most sophisticated deployments of gen AI in insurance globally.
Life insurers typically have low customer engagement compared with other customer-facing industries, which makes it challenging to get in front of customers to update information, create stickiness, and increase share of wallet. The same insurer built a solution using developed insurers’ most common existing analytical AI models (leads engines, next-best-product and next-best-action solutions, underwriting tools, and lapse propensity) by adding solutions mastered by teledirect insurers (optimal channel, day of the week, time of day, and tone of voice) and built a “language” and engagement layer to generate output across different media (text, in-app, email, call-centre scripts, and adviser tools).
The trickiest part was to avoid crossing the line into automated financial advice; with the “agent in the loop”, this is controlled. However, checking for bias, hallucination, regulatory compliance, and appropriate style was still a critical part of the development.
Another insurer in South Africa is using gen AI similarly to develop personalised and gamified educational content to help with self-led financial planning.
Moreover, a number of South African life and non-life insurers are using gen AI to automate, enable, and standardise customer underwriting, servicing, and claims operations, even with complex cases.