This article was produced in partnership with FM
Insurer FM has quietly built one of the world's largest risk prediction systems, identifying facilities 30 times more likely to suffer catastrophic losses.
Artificial intelligence has become a familiar presence in boardroom presentations and technology conferences. Yet while many companies chase the latest AI trends, FM has spent more than a decade building something far more substantial: a predictive system that can identify which industrial facilities are heading for disaster.
"At FM, we've been harnessing the power of AI for more than a decade – an eternity in the tech world – developing predictive tools that leverage conventional AI to understand and quantify risks," said James Thompson, senior vice president Asia Pacific division manager at FM. This approach reflects FM's methodical strategy of substance over hype, drawing on its history of testing sprinkler systems and fire walls since the 1880s.
The insurer's commitment to data-driven risk assessment has produced remarkable results. "Over the past 15 years, we've captured more than one billion data points about risk characteristics of the locations we insure," Thompson noted. This vast repository forms the foundation of their AI capabilities, distinguishing them from competitors who may lack such comprehensive historical data.
FM's latest AI tools represent a significant evolution beyond basic risk modelling. "Our latest AI tools incorporate interactions between hazards at a facility, substantially improving predictive power," Thompson said. This sophisticated approach recognises that industrial risks rarely occur in isolation – a mechanical failure might trigger an electrical fault, which could lead to a fire.
The results speak to the power of this integrated approach. "Our advanced predictive analytics can now identify facilities that are more than 30 times as likely to experience a severe loss as others, with the severity of those losses probably being more than 100 times greater than what we see from best-in-class facilities," Thompson said.
These predictions enable the insurer to deploy resources with surgical precision. Rather than spreading field engineers thinly across all clients, FM can focus expertise where it will have the greatest impact on preventing catastrophic losses.
The practical applications prove most valuable in sectors where industrial accidents carry enormous consequences. Power generation, mining, and chemical processing facilities face elevated risks from fire, mechanical breakdown, and electrical failures – scenarios where FM's predictive capabilities deliver substantial value.
"AI enables us to optimise our risk improvement efforts, helping our clients achieve more than twice the impact with the same capital expenditure," Thompson said. For industries operating on tight margins, this efficiency gain can mean the difference between profitable operations and financial disaster.
The human element remains central to FM's strategy, even as AI capabilities expand. "While AI plays a vital role in our operations, we never lose sight of the invaluable experience and judgement of our field engineers," Thompson emphasised. This balanced approach combines computational power with professional expertise, ensuring that predictions translate into practical risk mitigation strategies.
Climate change and globalisation have introduced new categories of risk that traditional insurance models struggle to address. "As secondary perils like floods and supply chain disruptions become more prevalent, FM is uniquely positioned to help our clients address these challenges," Thompson said.
FM's response includes developing comprehensive solutions for flood mitigation, roof securement, and supply chain vulnerabilities – all supported by AI-powered analysis. This expansion reflects the insurer's recognition that modern industrial risks extend far beyond the factory floor.
The company's climate resilience product suite demonstrates this broader thinking, incorporating tools like the Climate Resilience Tracker, the 2025 FM Resilience Index, and its Natural Hazard Toolkit, providing clients with essential solutions for growing climate risks.
These resources help clients understand and prepare for risks that may not have existed when their facilities were originally designed.
FM's approach to emerging AI technologies maintains the same measured philosophy that has guided their decade-long AI journey. "Looking ahead, we're excited to integrate generative AI functionality and a consistent usage strategy across our enterprise, always with a human in the loop," Thompson said.
This cautious optimism reflects hard-won experience about the gap between AI promises and practical applications. "Our measured approach ensures that these advancements are robust and secure, maintaining the integrity of our solutions," he noted.
The insurer's commitment extends beyond technology adoption to fundamental business strategy. "At FM, AI is not just a trend; it's a core component of our strategy to secure a productive future for our clients," he said.
For an industry built on managing uncertainty, FM's billion-data-point foundation provides a rare competitive advantage. As industrial risks become more complex and interconnected, the company's long-term investment in AI-powered prediction appears increasingly prescient.
