Intelligent AI has launched its report ‘The Risk Intelligence Gap’ based on analysis of more than 30 insurer annual reports and 20 interviews with senior executives across carriers, brokers and advisory firms, including Zurich, Generali, AXA, WTW, Hiscox and Tokio Marine Kiln.
The research data reveals an industry operating with limited visibility at the point of decision. In the UK, 93% of properties are insured for the wrong amount, while only 7% are adequately characterised in underwriting files. At the same time, 40-50% of submissions arrive incomplete, forcing underwriters to spend more than half their time chasing missing information rather than assessing risk.
The implications are material, as there is a fundamental asymmetry in underwriting economics; selecting the wrong risk has far greater consequences than mispricing it. This dynamic is driving adverse selection across portfolios, where insurers lacking high-quality data are more likely to retain mischaracterised risks.
The report identifies a major asymmetry in underwriting economics: interviewees argued that selecting the wrong risk can be exponentially more damaging than marginal pricing error. One senior underwriter summarised the imbalance bluntly: “Pick the wrong risk, and it could cost you a hundred million. Wrong pricing, maybe a hundred thousand.”
“Those insurers that address the data problem will be far better positioned to select risk, compete for higher-quality business and accelerate agentic AI underwriting, while those that don’t will face increasing exposure to volatility and adverse selection.
“The good news is that the data already exists, however, it isn’t reaching decision-makers in a structured, trusted form. Closing that gap is critical to unlocking the full value of AI and ensuring underwriting decisions are grounded in reality, not assumption.” said Anthony Peake, Founder and CEO, Intelligent AI
The research also points to measurable performance gains from better-quality exposure intelligence. McKinsey estimates that insurers using advanced analytics in underwriting have achieved 3–5 point improvements in loss ratios alongside 10–15% increases in new business. The report additionally cites examples of carriers reducing loss frequency by as much as 60% through the use of verified property-level risk data.
The impact is already visible. UK property claims reached record highs of £5.7 billion in 2024 and £6.1 billion in 2025, while global catastrophe losses continue to rise. In the case of Business Interruption, it is significantly under-priced, often by a factor of three to five, due to a lack of visibility into supply chain dependencies.
Across the insurance ecosystem, data exists within brokers, third-party providers and internal systems, but rarely arrives in a consistent, trusted and decision-ready format, despite significant investment in AI and automation. This creates a bottleneck at the submission stage, driving delays, manual intervention and weaker decisions.
In response to these industry challenges, Intelligent AI recently introduced Risk_API, designed to help insurers enrich underwriting workflows with verified, structured property intelligence at the point of submission. The company sees this type of API-first risk infrastructure as foundational to the next generation of AI-driven and agentic underwriting, helping carriers move from fragmented, manually gathered information toward continuously accessible, decision-ready risk intelligence embedded directly within underwriting operations. Continues…
For senior leaders, the implications are strategic. Rising regulatory expectations around data governance and auditability, combined with reinsurance pricing pressures, mean data quality, and its use with AI tools, is now directly linked to capital performance. This means that the ‘Risk Intelligence Gap’ is no longer an operational issue; it is a determinant of competitiveness and resilience for the future.
|
