Sapiens Insurance AI Maturity Index: 67% expect AI autonomy by 2029

Sapiens International Corporation has released its first Insurance AI Maturity Index, examining how insurers deploy artificial intelligence, prepare for agentic AI and manage the organisational changes associated with adoption.

The international study, conducted by Research in Finance, surveyed senior insurance professionals across companies of different sizes. Sapiens presented the findings at its Ignite 2026 event on 7 October in London.

According to the research, 67% of respondents expect AI in insurance to operate with full autonomy within the next two to three years. The prediction comes amid a considerable gap between insurers’ ambitions and their current technology capabilities.

Although 87% of insurers prioritise agentic AI, only 16% have adopted the technology. Conventional AI is further established, with insurers reporting measurable improvements in product deployment, operational costs and customer service. Among respondents, 77% report AI reducing product deployment times by an average of three months.

Only 25% of insurers reach the highest AI maturity category

The Sapiens Insurance AI Maturity Index assesses insurers against four areas of organisational readiness: cloud adoption, master data strategy, AI governance and workforce policies.

Companies fall into four maturity categories, ranging from Reactive organisations at an early stage of adoption to Autonomous Leaders with more developed technology and management capabilities.

AI maturity category	Share of insurers	Characteristics
Autonomous Leaders	25%	Cloud and some SaaS adoption, established master data and HR strategies, AI governance policies
Operational	28%	Cloud and some SaaS adoption, workforce initiatives and governance policies planned or implemented
Enabled	23%	Cloud adoption without SaaS, early master data and governance work, workforce initiatives being planned
Reactive	24%	Early cloud and data strategy development, initial governance policies and some HR initiatives

The distribution indicates considerable differences in insurers’ preparation for more autonomous technology. Three-quarters of surveyed companies remain outside the highest maturity category, despite widespread expectations of rapid AI adoption.

Paul Wheeler, CEO of Sapiens, said insurers face increasing pressure to accelerate their AI programmes. He identified agentic AI as an opportunity to improve customer relationships, shorten product launch cycles and deliver better digital experiences.

The research also indicates companies need suitable governance, workforce policies and data management practices before expanding autonomous AI applications.

Benelux and Nordic insurers lead AI maturity rankings

Regional results show Benelux insurers have the highest proportion of Autonomous Leaders at 33%, followed by the Nordics at 28%. The United States records 25%, compared with 23% in the UK and 18% in South Africa.

Property and casualty insurers are further advanced than life insurers. Some 30% of P&C companies qualify as Autonomous Leaders, against 25% of life insurers.

Company size also influences AI maturity. Among global insurers employing more than 20,000 people, 39% reach the highest category.

The proportion falls to 27% among large insurers with up to 5,000 employees and 21% among specialist or mid-sized companies employing fewer than 1,000 people.

These differences suggest larger insurance groups have progressed further in establishing the technical infrastructure and internal policies required for AI deployment.

Regional adoption patterns aren’t uniform. Nordic insurers demonstrate stronger interest in agentic AI and greater readiness to introduce it, whereas the UK ranks lowest among surveyed markets on both measures.

Agentic AI adoption trails conventional automation

Across the international sample, 61% of insurers identify AI as an important future business strategy, while 51% expect the technology to provide a competitive advantage. Actual deployment tells a different story.

Respondents report average usage levels of 40% for automation and 42% for AI. Agentic AI stands at 24%, indicating a slower transition towards systems capable of carrying out tasks with greater independence.

Autonomous Leaders are further advanced in agentic AI adoption. Among Reactive insurers, only 16% report using the technology.

The distinction matters because conventional automation and AI already deliver measurable operational improvements. Agentic AI has yet to demonstrate benefits at comparable levels across the surveyed organisations.

AI delivers stronger operational benefits than agentic AI

The study compares reported improvements across four insurance operating measures.

Reported benefit	AI	Agentic AI
Faster turnaround times	60%	44%
Improved customer experience	58%	47%
Lower manual handling costs	53%	44%
Fewer errors	51%	47%

Traditional AI records stronger results in every category. The largest difference concerns turnaround times, where 60% report benefits from AI compared with 44% from agentic AI. Customer experience follows a similar pattern, with 58% reporting improvements through AI and 47% through agentic systems.

Cost reductions show a nine-percentage-point difference. Error reduction produces the narrowest gap, with AI at 51% and agentic AI at 47%.

The findings suggest established AI applications are delivering more consistent operational results at present. Agentic AI adoption remains less widespread, limiting the extent of reported benefits.

Human oversight remains common despite high confidence in AI

Insurers express considerable confidence in automated decisions, although human supervision remains widespread. Among organisations using AI or agentic AI, 86% report being very or somewhat confident in decisions produced by these systems.

Yet humans still review an average of 53% of agentic AI work. Across AI and agentic AI applications, employees override approximately 15% of automated decisions.

Regional findings reveal differences between confidence levels and intervention rates.

The Nordics and South Africa each record 89% confidence in their checks on agentic AI. Corresponding figures are 78% for North America and 74% for the UK.

In South Africa, 48% of respondents report overriding more than 20% of AI decisions. The proportion reaches 35% in the Nordics, compared with 32% in North America and 29% in the UK. These results indicate insurers continue to rely on human review even where confidence in AI decision-making is relatively high.

The frequency of intervention also raises practical questions about how quickly organisations will move towards the fully autonomous operations anticipated by many respondents.

Workforce training and AI governance shape adoption readiness

For insurers that haven’t adopted AI, the research identifies governance, employee skills and data strategy as the main areas requiring additional support.

Workforce preparation differs considerably between maturity categories. Among Autonomous Leaders, 71% prioritise employee upskilling, compared with the international average of 53%. Only 35% of Reactive insurers report the same emphasis.

Some 57% of Autonomous Leaders are accelerating recruitment of employees with AI expertise. The international average is 38%, falling to 20% among Reactive insurers.

Nordic insurers also report more extensive requirements for employees to use AI. Across the region, 49% mandate AI use for all employees, compared with 25% in the UK and 17% in North America. These workforce differences accompany the regional variations in agentic AI readiness identified elsewhere in the Index.

For insurance companies planning further investment, the findings place employee training and governance alongside technology deployment as areas requiring attention.

Insurance AI adoption faces a gap between expectations and deployment

The 2026 Sapiens Insurance AI Maturity Index presents an industry with ambitious expectations for autonomous systems but uneven adoption across markets and business segments.

Two-thirds of respondents anticipate full AI autonomy within three years, although only one-quarter of insurers currently qualify as Autonomous Leaders.

Agentic AI remains less established than conventional AI, and human intervention continues across more than half of agentic AI work.

The research also documents tangible gains from existing AI applications, including shorter product deployment periods and improvements in processing times, customer experience and operating costs.

Insurers with more developed cloud infrastructure, governance policies and workforce programmes occupy the highest maturity category. Smaller companies and less advanced organisations report lower levels of readiness.