"The next generation of underwriting" – analytics officer on AI and its role in closing insurance

"The key to success lies in getting the balance right"

"The next generation of underwriting" – analytics officer on AI and its role in closing insurance

Life & Health

By Kenneth Araullo

When people are asked to point towards where the heart of insurance’s digital transformation is, most would look towards the clear market leaders in the United States or London, the birthplace of modern insurance. For some, however, it’s in the East; Munich Re APAC, Middle East and Africa life and health chief analytics officer Lee Sarkin (pictured) placed Asia at the “fore of digital transformation” in the industry.

In conversation with Insurance Business Asia, Sarkin spoke about the firm’s AI-powered underwriting solution, a proposition that was launched first for the region rather than the Western markets. Sarkin said that it makes sense, given Munich Re’s strong foundations in Asia.

“The rate of digitalisation has rapidly accelerated during the pandemic. Notably, automated underwriting rule engines (UREs) have been the starting point for life insurers to improve the overall customer experience and operational efficiency,” Sarkin said. “Munich Re’s URE, Allfinanz, is the clear global leader, used by over 130 insurers, and is the most widely used URE by insurers in Asia. Over many years, our URE paved the way for insurers to digitise the underwriting process. In the process, a critical byproduct was generated: good-quality data that is highly predictive of final (human) underwriting decisions.”

Given the fact that Asia had a large base of insurers using this URE, he explained that it’s only natural for the firm to launch in the region first. With Munich Re’s knowledge of the data and the software itself, Sarkin said that the firm has now developed robust AI models that should help insurers monetise the data and transform their customer experiences.

“We’ve significantly uplifted underwriting automation (straight-through processing) for multinationals (MNCs) through our next-generation AI-augmented underwriting solution. This was well-timed since it addressed the priorities of Asian insurers whose digital strategies focus on improving the customer experience via data, AI and tech,” he said.

Sarkin also said that the firm is leading the way for an underwriting proposition through digitisation and automation via UREs.

“Now again, we are enabling insurers to transition to the next generation of underwriting by packaging data, AI models, UREs and machine learning operations (MLOps) technology such as our Predictor AI platform,” Sarkin said.

Closing the insurance gap, and the importance or “risk guardrails”

Sarkin noted the fact that AI spending in the region – including software, services, and hardware of AI-centric systems – will reach $49.2 billion in 2026. This figure, he said, highlights the immense potential of the technology for the market, and not just insurance.

“To this end, we are confident that AI and automated solutions will continue to rank high on the agenda for insurers and offer a range of benefits that improve the customer experience, operational efficiency, risk selection, and ultimately sales and profitability,” he said.

With these in mind, Sarkin is convinced that AI-augmented underwriting solutions are key to breaching the insurance gap in the region, primarily by making it easier and more convenient for customers to purchase insurance, while at the same time reducing distribution and administration costs for insurers.

Unrestrained proliferation does come with its own set of risks, as an AI-focused company CEO already highlighted. In support of this, Sarkin said that a lot of the risks come down to how AI is tempered with the human touch of the industry.

“We have successfully augmented the traditional underwriting process with AI and in the process, significantly upskilled underwriters and actuaries with new AI skills. Domain experts like underwriters and actuaries play a key role in the success of AI-driven solution developments,” he said.

Munich Re’s own augmented underwriting solution, Sarkin said, is a true hybrid of three – AI, underwriting rule engines (URE), and expert underwriters. Central to this collaboration are “risk guardrails” that leverage the strength of all three while covering the others’ weaknesses.

“For instance, signing off the risk of AI models will continue to rely on actuaries as a human element to manage the risk of AI model errors. Furthermore, future data to retrain AI models also require a sample of the traditional underwriting decision by human actuaries. Another example of future augmentation is the AI-assisted underwriter who can responsibly leverage the strengths of large language models,” he said.

“The key to success lies in getting the balance right”

Sarkin places great importance on the role of responsibility in developing AI systems like these, and to that end he said that the firm is keeping up with all things related to AI regulation in the region.

“We [are] actively involved in AI regulator-driven industry consortiums, such as the Singapore Monetary Authority of Singapore’s (MAS) Veritas initiative on responsible AI,” Sarkin said. “We’ve developed responsible AI tools and methodologies that ensure AI-augmented underwriting models comply with responsible AI regulations. Given the risk complexity of AI-augmented underwriting, as well as the deep domain knowledge needed to manage these challenges, we have pioneered our AI risk implication methodology to enable insurers to manage and sign off the corresponding risks.”

There are also implications to giving an AI power over the customer experience; Sarkin said that there should be balance in improving the former and streamlining underwriting, as tipping the scales in the latter’s favour could result in incorrect decisions.

“The stakes are high, as AI-based underwriting errors can result in substantial underwriting losses,” he said. “The key to success lies in getting the balance right. This requires insurers to manage a new ‘risk complexity’ arising from the interplay of required capabilities including domain knowledge of UREs and their complex data, developing AI models that comply with responsible AI regulations, securing sign-off from actuarial risk management before model deployment on the risks posed to insurers’ future claims, integrating real-time AI model predictions into UREs and safely maintaining AI models in production via MLOps technology.”

Munich Re’s aim in the future is to be a global market leader for enabling clients to transition to the next generation of underwriting, one that’s powered by AI, data, and modern technology. Sarkin expressed his belief – supported by the firm’s new AI-powered proposition – that the firm is well-positioned to take a leading role in this growing market.

“We also walk this journey together with clients – our ability to accept risk on insurance policies underwriting by AI models enables us to offer alignment of interest and peace of mind to insurers. By forging long-term partnerships with Munich Re, insurers can benefit from existing capabilities to achieve speed-to-market and avoid a high total cost of ownership, allowing them to transition smoothly to the next generation of underwriting,” he said.

What are your thoughts on this story? Please feel free to share your comments below.

Related Stories

Keep up with the latest news and events

Join our mailing list, it’s free!