Structural risks: Existing challenges (and opportunities) for the insurance industry

Structural risks arise from fundamental trends shaping the economy, society and the environment, such as shifts in demographic and climate conditions. Managing these risks is crucial for insurers to protect policyholders, to safeguard their own operations and to support macroeconomic resilience. Here we highlight five structural risks that matter for the insurance industry today: declining trust in institutions, aging populations, social inflation, mortality risks and digitalisation.

A sense of unfairness can erode trust. As many as 61% of respondents to the 2025 Edelman Trust Barometer expressed a moderate to high sense of grievance against businesses and governments.¹ Forty percent went as far as to endorse hostile activism, including violence and damaging property, to drive change. The insurance sector has not been immune from grievance-driven trust deficit.

Surveys show that in most markets, trust is an important determinant in consumers insurance purchasing decisions. For instance, more than 80% of commercial buyers say trust is an important factor influencing the decision get their business insured.² Lack of consumer trust can impact insurers at many levels. Policyholders may seek to change providers or not buy insurance at all for certain risks. When there is distrust, for instance with consumers not convinced that insurers will pay out on claims submissions, or perhaps because of complex policy wording that they do not understand, it is harder for insurers to close sales.

Lack of trust could also entail reputational risks for insurers. Several recent surveys indicate that less than two thirds of consumers trust insurers. Specifically, a survey by the European Insurance and Occupational Pensions Authority (EIOPA) last year found that only 50% of consumers trust that insurers would pay out on claims for losses resulting from natural catastrophes.³ With insured losses from natural catastrophes rising by an annual long-term trend of 5-7% in inflation-adjusted terms, sustaining trust will require ongoing efforts.⁴

As a broad term, we define “social inflation” as encapsulating those factors leading to increased insurance claims severity beyond that explained by economic drivers. Over the past decade, such factors have accounted for 57% of the increase in liability claims in the US.⁵ Since 2020, the number of nuclear verdicts defined as awards over USD 10 million) in the US have more than quadrupled, while the median verdict value has more than doubled (from USD 21.5 million to USD 51 million).⁶ Due in part to higher litigation awards, US liability lines exposed to bodily injury claims saw profitability deteriorate over the five years to 2024, with cumulative underwriting losses of USD 43 billion.⁷ In 2024, the average award in cases against a corporate defendant in the US rose to USD 65.7 million, up from USD 41.7 million in 2023, according to data from LexisNexis, as reported by the Financial Times.⁸

Social inflation is difficult to measure and predict and can significantly reduce the overall risk transfer capacity of the insurance industry. It is particularly disruptive in liability insurance business because it affects longest-tail lines disproportionally, these typically being more exposed to legal system developments. Median limits purchased for liability towers⁹ declined by an average of nearly 25% in nominal terms and by 46% in inflationadjusted terms between 2014 and 2023, a period of increasing loss costs.¹⁰

Though to date mostly an US issue, we expect other regions will see an increase in the size of legal settlement awards within the next five years, particularly Europe, driven by easier access to litigation, expansion of collective redress and broadening of the product liability landscape. Third-party litigation funding (TPLF) could an important driver, more so in the absence of TPLF specific regulation.¹¹ The European Union supports and is promoting collective actions across various areas such as data privacy, environmental, social and governance (ESG), and product liability. The revised Product Liability Directive, for instance, eases the possibility of mass litigation by altering the burden of proof, particularly in complex digital and AI cases. While collective redress is still uncommon in many EU countries, it is rising steadily, driven by new regulation like the Digital Markets Act and General Data Protection Regulation (GDPR).¹²

In the US and UK, excess mortality from COVID-19 is expected to remain positive at least until 2027, and possibly until 2033. The degree of this positive excess mortality varies by scenario.¹³ Recent experience in the US shows that life expectancy rose by 0.9 years to 78.4 in 2023, still slightly below the pre-pandemic high of 78.8.¹⁴ Mortality rates in the UK, meanwhile, were reported to have returned to pre-pandemic levels in 2024.¹⁵ In the euro area, excess mortality stood at 3% at end 2024, with fluctuations through the year and between member states.¹⁶ Australia too has seen continued excess mortality since COVID-19, the latter accounting for 2.2% of deaths in 2023-24.¹⁷ Meanwhile, mortality rates in the UK were reported to have returned to pre-pandemic levels in 2024.¹⁸

Elevated levels of excess mortality are a potential challenge for L&H insurance, with potentially several years of elevated mortality claims ahead, depending on how general population trends translate into the insured population. Ongoing excess mortality can have implications for L&H claims and reserves. Excess mortality that continues to exceed expectations may affect the long-term performance of in-force life portfolios, and also the pricing of new life policies.

Factors other than COVID-19 drive variances in mortality trends across markets too. The US continues to diverge on its different mortality trend from other developed nations. Recent studies show rising “avoidable mortality” in the US (by 32.5 avoidable deaths per 100 000 people), but a decrease in the European Union (EU) and Organisation for Economic Co-operation and Development (OECD) countries (by 25.2 and 22.8 avoidable deaths per 100 000 persons, respectively).¹⁹ Last year a study highlighted how the US and the UK had quite different life expectancies at birth, mostly due to higher rates of preventable deaths in the former. These were mostly on account of higher rates of cardiovascular related deaths, followed by drug overdoses, firearmrelated homicides and suicides, and motor vehicle crashes.²⁰ The proportion of the causes of death differ by socioeconomic groups.²¹

Innovations like GLP-1 weight-loss medications and the use thereof are becoming more widespread. Medical advances that reduce obesity should help reduce mortality in the future. Long-term data is still emerging and early results are promising. However, studies indicate that many individuals regain a significant portion of the weight they have lost when they stop taking the medications.²² In the long term, positive mortality improvements will rely on advancements in cardiovascular disease and developments for the treatment of cancer. These remain a major causes of mortality and critical illness among the insured population.

The world’s ageing population is being driven by increases in life expectancy and declining birth rates. According to data from the UN, the share of persons aged 65 and above in high-income countries is forecast to trend up through to the mid-2040s. The number of 25 to 49 year-olds is expected to shrink 5% by 2050.²³ As life events like birth of a child can spur life insurance purchases, a population that is ageing implies lacklustre growth in demand for mortality protection and accumulation savings premiums. Fewer and later family formations magnify this headwind.

For the L&H industry, growing longevity risk pools (ie, ageing populations seeking health and retirement security) are a large premium opportunity. A 65-year-old stopping work today faces a 16% longer retirement than one who left the workforce in 2000, and nearly 40% more than someone who retired in 1975. This will continue as life expectancy post age 65 may rise to 20 years by 2050 (compared to 16.8 in 2020).²⁴ Using per capita medical expenses and proportion of population in age categories, we calculate that by 2050, the 65+ population in the US will account for almost 50% of health expenditure.^{25, 26} In Japan, the 65+ age group already accounts for more than 60% of total spending on healthcare services.²⁷

The evolution of digital technologies is reshaping the risk landscape and could prompt to more demand for first- and third-party liability covers. The number of AI incidents reported has been increasing. From 2023 to 2024, it grew by more than 60%.²⁸ A third of these incidents occurred due to AI systems failure. As more people and businesses engage with AI, there may be a corresponding increase in lawsuits.²⁹ Recent spikes in litigation have been based on allegations of intellectual property infringement and defamation lawsuits (some due to large language models such as ChatGPT). This is not a risk in isolation. The lawsuits could be contributing factors to social inflation in the future.

Newer approaches such as GenAI can also be vulnerable to sophisticated fraud like deepfakes, as well as heightened liability concerns.³⁰ Losses can be categorised into four broad risk categories: 1) fairness, bias & discrimination; 2) legal & regulatory compliance; 3) malicious use & safety; and 4) unintended outcomes & malfunctioning products. Over the years, the share of incidents due to fairness, bias and discrimination has declined, while those of malicious use have risen. The insurance industry is in initial stages of the associated product development cycle and clarity around coverage, exclusions and standardised wordings remains pending.