AI’s Emerging Threat Landscape for Insurers
The Lloyd’s Market Association (LMA) mid-2025 survey collected 144 responses, of which 94% were underwriting professionals, to map likely AI-related loss scenarios and their severity. The exercise responds to growing adoption of generative models and automation across financial and commercial services and aims to give underwriters a shared evidence base for emerging exposures.
Sector-Specific Insights on AI Loss
Professional Indemnity: Top Concern
Underwriters rated Professional Indemnity as the principal near-term risk. Widespread use of AI in advice, reporting and decision support raises the prospect of systemic errors, biased outputs or flawed model training producing professional losses. Claims could be large where AI-driven advice influences high-value transactions or fiduciary decisions.
Cyber Risk & System Downtime
Respondents see cyber exposures and AI-caused system downtime as plausible and potentially disruptive. Vulnerabilities include model tampering, supply chain compromise and failures in model orchestration that cascade into business interruption. Risk management and access controls are central to reducing frequency and impact.
Product Recall & Autonomous Vehicles
Product recall was judged currently low in likelihood but likely to grow as AI embeds into physical products. Opinions on Accident & Health and self-driving vehicle severity were mixed; most underwriters expect exposures to rise over time as deployment scales and liability attribution becomes clearer.
Preparing for AI’s Underwriting Impact
“We expect AI use to accelerate across the market, and with that the potential for new and larger exposures,” said David Powell, LMA Head of Technical Underwriting. The survey signals priorities for action: update policy wordings, adopt scenario testing, require model governance disclosures, improve vendor oversight and coordinate cross-line aggregation analysis. For insurers and investors, the immediate steps are structured assessment, clearer underwriting questions and active participation in market-wide standards to price and limit accumulating exposures.
Key takeaway: AI is reshaping loss patterns across multiple lines. Underwriters who build practical governance checks, clarify coverage boundaries and stress-test portfolios will be better positioned as claims emerge and technologies evolve.
