Cautious Steps Towards Artificial Intelligence in Insurance

Despite surging global enthusiasm for artificial intelligence (AI), the insurance sector remains notably cautious in embracing the technology at scale. A recent report by the research and analytics firm IDC reveals that while most insurers recognise AI’s strategic potential, only a small minority have progressed beyond experimentation to meaningful, enterprise-wide deployment. The study, commissioned by analytics company SAS, suggests that structural, cultural and governance-related constraints continue to slow adoption across the industry.

According to the report, titled The Impact of Data and Artificial Intelligence: The Inevitability of Trust, just 7 per cent of insurance companies describe their AI capabilities as “transformational”. These organisations have managed to embed AI deeply into core business processes such as underwriting, claims management and risk assessment. By contrast, 14 per cent of insurers still rely on fragmented and siloed data infrastructures, which significantly hinder innovation, automation and data-driven decision-making.

Although the use of AI within insurance is gradually increasing, the report highlights a persistent gap between ambition and readiness. Many firms lack mature data management practices, robust governance frameworks and a culture of organisational trust—factors that are essential for deploying AI responsibly and at scale. Without these foundations, insurers struggle to align technological initiatives with broader business objectives.

Trust emerges as a central theme of the research. Survey respondents indicated that they tend to place greater confidence in generative AI tools than in traditional AI systems. However, this confidence is often not supported by rigorous governance, risk controls or high-quality data management. As a result, insurers face a dual risk: on one hand, scepticism prevents them from fully leveraging proven AI solutions; on the other, excessive reliance on emerging technologies without sufficient oversight may expose them to operational, ethical and regulatory vulnerabilities.

Investment patterns further reflect this cautious stance. Only 8 per cent of insurers plan to increase AI-related spending by 20 per cent or more over the next year. Around 60 per cent expect to raise investment modestly, within a range of 4 to 20 per cent, while nearly one-third anticipate minimal growth—3 per cent or less—or even a reduction in spending. This indicates that for most companies, AI remains largely confined to pilot projects rather than full-scale implementation.

The disparity between confidence and capability is also pronounced. Merely 9 per cent of insurers report both high trust in AI and the internal capacity to implement it effectively. In contrast, more than 40 per cent occupy a position of imbalance, where either trust or capability is lacking. This misalignment points to weaknesses in governance, skills development and ongoing oversight.

Data-related challenges stand out as the most significant obstacles. More than half of surveyed firms cite poor data governance and fragmented data architectures as major barriers, while 44 per cent highlight a shortage of skilled AI professionals. IDC Research Director Cathy Lange notes that, compared with other industries, the insurance sector demonstrates one of the lowest levels of AI maturity, limiting its ability to scale innovation.

Selected Indicators from the Report

The report concludes that the insurance industry stands at a critical juncture. While AI adoption is advancing incrementally, failure to improve data quality, governance structures and workforce capabilities could leave insurers increasingly uncompetitive compared with more digitally mature sectors.