How Asia’s insurers are using AI — and where it could take them

Insurance leaders respond with a clear ambition, the technological and operational capability to deliver, and a fast path to adopt changes.

The pressures on Asian insurers have been building for years, from shifting demographics to rising climate-related losses and a large, underinsured population. None of these problems are new, and no single technology will solve them. But artificial intelligence (AI) is giving insurers a genuine opportunity to make progress on all of them, and to run their businesses better along the way.

Across the region, AI is moving from experiment to practice. What matters now is how fast, and how purposefully, each market, and each company makes that shift. The differences are already visible, and they will shape the industry’s competitive landscape for years to come.

Consider what is underway. Over the past two decades, whilst insurers globally are struggling to turn technology investment into lower operational costs — 17% higher globally and 60% for North America — Asia Pacific is the only region that managed to cut the cost ratio by 27%.

The improved efficiency is a key commercial incentive for the next wave of investment. Property and casualty insurers in Asia-Pacific are expected to increase technology spending by nearly 14% in 2026, a particularly strong rate of growth even as IT spending accelerates worldwide.

AI is an increasingly significant factor in those decisions, and each of the region’s major markets is putting it to work in its own way.

In China, health insurers are using predictive analytics to manage chronic disease risk, identifying customers likely to develop conditions such as diabetes or cardiovascular disease and adjusting products accordingly.

In Japan, some property insurers are adopting computer vision to assess damage claims. Japan’s Financial Services Agency has published voluntary guidance encouraging risk management and human oversight for AI use.

India has become a testing ground for algorithmic underwriting, compressing the time it takes to bring new pricing models to market.

And in Singapore, insurers operate under the Monetary Authority of Singapore’s proposed AI Risk Management Guidelines, which recommends AI outputs used in customer-facing decisions be explainable.

The regulatory detail in that list is worth pausing on. In Japan and Singapore, supervisors are already writing rules for the technology rather than waiting to see how it develops. For insurers, this means the rules of the road are being written whilst they drive, and the companies shaping them are the ones already deploying the technology.

The results are already measurable. Business case reviews of brokers find that firms using AI to engage clients more systematically — reaching them with the right message at the right point in their journey — have cut churn by upwards of 50%. Amongst specialist underwriters, risk engineering tools have cut quoting times from more than a month to a matter of days.

Gains of this kind compound. A company that shortens its quoting cycle wins more business, which generates more data, which improves its models, which shortens the cycle further. The same logic applies to customer retention, claims accuracy, and fraud losses. The financial consequences are already measurable.

A recent study finds that over the past five years, insurers leading on AI have delivered 6.1 times the total shareholder return of those lagging behind. In most other sectors, the gap between AI leaders and followers runs to two or three times. Insurance is pulling apart faster than most industries, and the pressure will come from digital-native entrants as well as established peers.

How should insurance leaders respond? A rewired enterprise approach with a clear ambition, the technological, and operational capability to deliver, and a fast, staged path to adopt changes across the organisation. Under this strategic approach, four priorities stand out.

First, take an honest measure of where your company stands.

Virtually every insurer now has AI use cases in production. Far fewer have moved past pilots to capability that changes how the business actually runs, and many executives know it. The relevant benchmark is deployed capability that changes how the business runs, measured against peers, and against the new entrants that would be happy to poach your customers.

Second, build a consistent internal playbook.

Companies that succeed with AI tend to standardise their use cases, apply shared metrics across business units, and put clear governance around how models are built, validated, and monitored. Ad hoc adoption produces ad hoc results.

Third, plan for more than one future.

AI adoption will move at different speeds across markets, lines of business, and customer segments, shaped by regulation, data availability, and customer trust. Leaders should prepare for several scenarios rather than betting on a single adoption curve.

Fourth, prepare your people.

The skills the industry needs are shifting, and the shift will reward companies that invest in training and reorganising work before the pressure arrives, rather than after.

The ongoing pressures on Asian insurers will keep building whether or not the industry acts. What has changed is that the tools now exist to address them, and the region already has examples of what those tools can do. The insurers that learn fastest from them will be the ones setting the pace.