Insurance premiums are rising faster than inflation, putting increasing pressure on the affordability and long-term sustainability of cover. In France, the issue is becoming particularly visible as climate-related losses, higher claims costs, expensive repairs, reinsurance pressures and emerging cyber and geopolitical risks converge.
Earnix’s latest insurance trends research highlights the technology challenge behind this shift, with 95% of French insurers reporting that they have adopted artificial intelligence (AI), while just 32% say they trust their underlying data. The gap illustrates a wider problem for insurers: as risks become more complex, better technology alone will not be enough if the data supporting pricing and underwriting decisions cannot be relied upon.
Home insurance is among the lines facing the greatest pressure. Premiums increased by close to 10% in 2025 and are facing further upward pressure from worsening natural catastrophe losses and an increase in France’s Cat Nat surcharge from 12% to 20%. The levy is intended to replenish reserves within the country’s national natural catastrophe scheme.
The scale of the underlying risk is becoming harder to ignore. Between 1982 and 2024, cumulative Cat Nat claims in France reached €61.2bn. Research from CCR, based on Meteo-France data, estimates climate change could increase natural catastrophe costs by 40% by 2050, potentially rising to 60% when higher asset values and greater territorial exposure are taken into account.
But the problem extends beyond whether a risk can technically be insured. The bigger question is whether customers will continue to accept the price of that protection.
French Institute of Actuaries president Simon Le Dily has argued that insurance can remain technically viable while becoming economically or socially unworkable if premiums, deductibles or coverage gaps move beyond what policyholders are prepared to accept.
Signs of that pressure are already emerging. A survey of 368 insurance professionals by Transformers de la relation client found that 57.2% identified transparency around pricing and underwriting decisions as the most important factor in customer relationships. This was followed by alignment between price and compensation at 43.2% and price competitiveness at 42%.
That makes affordability and transparency increasingly intertwined. If insurers need to charge more to reflect growing risk, they also need to demonstrate why prices are changing and what customers are receiving in return.
The sustainability challenge is therefore becoming structural rather than cyclical. ACPR stress tests indicate that the French insurance sector can remain solvent through to 2050, but this assumes substantial further increases in premiums.
That creates a potential feedback loop. Higher prices can push some customers out of the market, reducing the size of risk pools. Smaller pools can leave remaining policyholders facing greater costs, while increasingly concentrated exposure can make certain risks harder to insure in the first place.
Technology is being positioned as part of the response. AI can help insurers model increasingly complex risks, analyse large datasets and identify opportunities for more targeted prevention. According to Earnix’s research, 33% of insurance executives are using AI to simulate risk scenarios, while 38% are using it to personalise customer recommendations.
Yet the relatively low level of trust in underlying data highlights the limits of simply adding AI to existing processes. If insurers are to use increasingly sophisticated models to price risk, identify vulnerable customers or recommend preventative action, the quality and governance of the data feeding those models becomes critical.
The French insurance system also continues to rely on a balance between private-sector pricing and state support. CCR chief executive Édouard Vieillefond has described CCR’s role as a systemic shock absorber when risks reach the limits of private insurability.
That model has helped France maintain broader insurance availability than some markets where insurers have withdrawn from high-risk areas or introduced wider coverage exclusions. But its long-term resilience will depend on better risk mapping, more personalised prevention and more effective use of data.
The urgency is underscored by the scale of recent climate losses. Wildfires in France in 2026 burned almost 100,000 hectares and generated more than €500m in losses, according to the figures cited in the source analysis.
For insurers, the challenge is therefore no longer simply how to price risk. It is how to keep insurance affordable and trusted while the risks being covered become more expensive and difficult to predict.
Earnix’s research points towards AI, scenario modelling and personalised recommendations as some of the tools insurers are using to respond, but the industry’s ability to make those technologies work will depend on having reliable data and maintaining customer trust. As climate risk continues to reshape insurance economics, the future of the model may depend on whether insurers can combine more accurate pricing with stronger prevention and clearer explanations of why customers are paying more.
Matmut Group director of information systems transformation François-Xavier Enderle said, “Customer data sovereignty is the backbone of the insurance model.” That principle could become increasingly important as insurers rely on more data-driven approaches to navigate the growing gap between what risks cost to cover and what customers can afford to pay.
Copyright © 2026 FinTech Global
