By Alex Acworth, insurance lead, Accenture

For the past few years, much of the conversation about AI in insurance has focused on what the technology can automate. But as adoption grows, insurers need to decide how that freed-up time should be used. 

The productivity gains are already becoming tangible. Accenture research found that 81% of insurers have achieved at least a 5% improvement in gross written premiums from data and AI initiatives. At the same time, 86% of insurance employees say AI tools have increased their overall productivity.

For insurers, the real benefit comes from redirecting that time towards work that improves the business and customer experience.

Insurance depends heavily on specialist expertise. Underwriters need to understand increasingly complex risks. Claims professionals need to interpret information and support customers, often at difficult moments. Brokers and distribution teams need to build relationships and understand the needs of increasingly sophisticated customers.

Yet highly skilled people still spend a significant proportion of their working day on tasks that do not make the best use of that expertise.

Accenture research into underwriting, for example, has found that around 40% of an average underwriter's time is spent on administrative and other non-core tasks. AI can help reduce that burden by extracting information from submission documents, bringing together relevant data and presenting it to an underwriter in a usable format.

Processing a submission faster is only part of the benefit. It also gives underwriters more capacity to assess complex risks, work with brokers and customers, and pursue opportunities that might otherwise be missed.

Turning productivity into customer value

Claims and customer service stand to benefit in much the same way.

In claims, AI can help gather and summarise information, surface relevant policy information and support employees managing workflows.

In customer service, it can help resolve routine enquiries and give employees faster access to the information they need. That gives employees more time to handle the complex or sensitive interactions that need that human touch.

This is particularly relevant in Australia, where customer communication, support for vulnerable customers and appropriate claims-handling resources remain important priorities for insurers and regulators.

AI could help insurers give their employees more time for this work; if technology can reduce the administrative workload surrounding a claim, for example, claims professionals can spend more time communicating with customers, managing complex cases and applying their experience where human judgment matters.

There is a commercial benefit to getting those interactions right too. Accenture research found that 47% of customers dissatisfied with how their claim was handled were considering switching insurers.

Insurers should therefore measure productivity by more than cost savings. If AI gives a claims professional more time to communicate with a customer, an underwriter greater capacity to examine a complex risk or a service employee the information required to resolve an enquiry more effectively, that productivity can translate into stronger customer relationships and better business outcomes.

The next challenge is scaling what works

Many insurers have already shown that individual AI tools can deliver value. The harder task is embedding them across the business in a meaningful way.

Accenture research found just 23% of insurers have achieved enterprise-wide AI integration. At the same time, 83% report moderate-to-severe gaps in the skills required to translate between AI capabilities and business needs.

This suggests the next phase should not be about deploying AI everywhere simply because the technology is available.

Instead, insurers should begin by identifying the specific problem they want to solve first: identifying where employees lose time gathering information, where fragmented systems cause customer delays, and where manual processes prevent people from making timely decisions or pursuing new opportunities. 

By addressing the issue at its heart, insurers can shift the focus from automating individual tasks to revolutionising how work gets done across the business.

The people best placed to identify valuable applications of AI are often those who understand the work most deeply. Underwriters, claims professionals, actuaries and distribution teams need enough understanding of AI to identify where it can improve their workflows, challenge its outputs and determine where human judgement remains essential.

That becomes even more important as agentic AI develops. Accenture research found 68% of insurance leaders believe integrating AI agents into core workflows will transform roles.

But insurers should not judge progress by the number of processes they automate. Instead, it should be whether AI helps the organisation make better decisions, respond to customers more effectively, pursue more opportunities and make better use of the expertise already inside the business.

Ultimately, the real productivity gain from AI will come when skilled employees spend less time on admin and more time making decisions, solving problems and supporting customers.