Global insurance firms are steadily increasing their capital expenditure on artificial intelligence (AI) with the primary objectives of reducing claims expenditures and refining underwriting processes. However, a significant proportion of these institutions continue to face difficulties in generating meaningful financial returns. Industry research indicates that these shortfalls are predominantly caused by obsolete legacy IT architecture and fragmented data management practices across the sector.
Projected Investment and Financial Outcomes
According to a report published in March by Boston Consulting Group, Inc. (BCG), capital allocation toward AI technologies within the property and casualty (P&C) insurance segment is projected to experience a substantial increase. Investment levels are forecast to more than triple, rising from previous baselines to reach 1.9% of total corporate revenue in 2026.
The consultancy’s analysis suggests that the depth of technology integration directly correlates with financial performance. Insurers that successfully implement AI deeply and systematically across their comprehensive operational frameworks could see operational costs decrease by approximately 20%. Concurrently, these deeply integrated systems have the potential to drive an increase in gross written premiums of up to 5%.
Current Applications and Structural Bottlenecks
At present, insurance providers are deploying AI tools within underwriting and claims processing departments. These applications are designed to automate repetitive administrative tasks, enhance structural pricing accuracy, and facilitate the early detection of fraudulent claims.
Despite these functional deployments, BCG noted that the majority of insurers utilise AI within isolated business units rather than across the entire enterprise. This siloed approach severely curtails the aggregate financial impact of the technology.
Data compiled from industry reports outlines the current operational impacts and performance benchmarks reported by insurance providers currently utilising AI infrastructure:
| Metric Category | Operational Performance Indicator | Reported Percentage of Insurers |
| Productivity Impact | Achieved modest improvements in workplace productivity | 63% |
| Productivity Impact | Realised measurable, significant gains in output per employee | 11% |
| Strategic Horizon | Expect AI to fundamentally reshape existing business models (2027–2029) | Near 60% |
| Future Funding | Planning further increases in AI capital expenditure (2026–2028) | 66.7% (Two-thirds) |
Technical Obstacles and Future Outlook
A separate research report published in April by A.M. Best Company, Inc. highlighted the persistent infrastructural barriers preventing optimal technology adoption. Industry experts emphasise that the core issue lies in the compatibility between modern algorithms and historical software.
Kaitlin Piasecki, an industry research analyst at A.M. Best, stated within the report:
“Legacy systems can create significant barriers when implementing AI because they simply were not built for this type of data integration.”
This sentiment regarding systemic fragmentation was echoed by Sridhar Manyem, Senior Director of Industry Research and Analytics at A.M. Best. Mr Manyem noted that AI applications consistently produce unreliable or sub-optimal outcomes when the underlying corporate data is fragmented or suffers from poor governance structures.
Nevertheless, corporate appetite for technological transition remains robust. A.M. Best’s findings confirm that two-thirds of surveyed insurers intend to expand their financial commitments to AI development between 2026 and 2028, underscoring a long-term industry commitment to overcoming these legacy hurdles.