But casualty lines present challenges as insurers and clients grapple with AI’s expected impact on coverage.
After years of high property-and-casualty insurance premiums globally, the rate reductions multinational companies saw in 2025 are anticipated to continue this year, especially for those willing to revisit their current carrier relationships.
“Nuclear” jury verdicts in corporate liability lawsuits will likely continue to push up casualty premiums in the US, and the trend is spreading to other countries. But these appear to be an outlier.
“With the exception of US casualty, for all other lines of business the starting point is double-digit percentage reductions,” between 10% and 20%, says Simon Delchar, global head of placement at Willis Towers Watson. And for companies willing to look beyond their current carriers, “the decrease could be significantly more than 20%.”
That’s in part because premiums have rested at high levels over the last half dozen years while insured losses, despite several major catastrophic events over that period, have remained relatively modest in aggregate. Consequently, global insurers’ profits have soared, as reflected in the stellar third-quarter earnings of giants such as AIG, Allianz, and Chubb, and many are releasing reserves from previous years. Reinsurers providing capital to primary insurers successfully renewed commitments at the end of 2025, further bolstering the industry.
With the wind at their backs, insurers are now anxious to grow their businesses.
“The impact is a big feeding frenzy,” says Delchar, putting further downward pressure on rates.
Global wildfire, flood, and typhoon events were milder than anticipated last year. That impacts the current pricing reductions, because weather is the primary driver of price.
That could quickly change, of course, notes Nadine Moore, managing director and senior partner at Boston Consulting Group. Even with recent reductions, “we’re nowhere near the trough we were in five years ago,” she says. “We can’t predict the weather, but there’s a view we’re going into a deep soft market.”
Some rates are tougher to budge. Significant rain in California in recent months reduces the risk of the wildfires that raged across the region a year ago, but negotiating lower premiums for wildfire coverage in the Golden State remains challenging. And certain industries, such as mineral extraction and energy, are also more difficult to cover, Delchar says.
US As An Outlier
Casualty lines present a more mixed picture. Rates were down overall in Europe and elsewhere outside the US in 2025, where at year-end they had increased by high single-digit percentages and as much as 15% for the most challenging risks.
“The real issue on US casualty is capacity,” Delchar says. “A year or two ago, markets were happily committing substantially more capacity than they are prepared to offer now, resulting in the need to go to more insurers to get the same capacity. And that’s more expensive.”
While rates vary state by state in the US, litigation and costly jury verdicts have pushed insurers to reduce their exposures and increase premiums, Moore says, and that’s unlikely to change soon. The actuarial models used to determine rates and what to reserve for the year ahead are based on historical performance.
“The casualty sector has seen many insurance companies update and increase their reserves, because the rate of inflation on claims is growing faster than in the past,” she says. “Not only is pricing changing, but companies have been forced to take bigger deductibles, so they’re retaining more risk. And those trends are still shaking out.”
The US is a particularly litigious society, but other jurisdictions with similar norms and legal systems, including Spain, Japan, France, Germany, the UK, and Australia, are heading in a similar direction, Moore says, noting that Italy’s different set of laws has so far thwarted the trend.
That could change as the EU’s Product Liability Directive comes into force by the end of this year. The PLD sets strict liability for damage caused by product defects, notes David Rahr, global leader of Marsh Multinational’s Multinational Client Service, meaning plaintiffs don’t necessarily have to prove negligence, but only that the product was defective and caused harm.
“That will be a big change and could bring US-style verdicts to other parts of the world, which presumably has insurance markets concerned,” he says.
Multinationals’ Advantages
A well-capitalized insurance industry is good news for companies, and especially multinationals whose heft gives them a strong hand in negotiations. Such companies typically rely on brokers to coordinate multiple carriers across different lines of insurance and jurisdictions. That broker, especially in a softening market, can use an assortment of techniques to negotiate lower premiums, such as offering larger positions to one or more carriers.
It’s not just the price of the premium that must be negotiated, however, but also the policy’s wording, Moore notes: “That’s very important, because I could get more coverage for the same amount of money.”
More accurately matching a company’s risks to the language of the insurance policy can also reduce premiums. Large, complex companies’ operational risks change quarterly if not more frequently, says Fred Barnachawy, CIO of risk management consultancy DeshCap, and it is imperative for them to track those changes. The smart move is to update policies as operational data changes, but most multinationals—and businesses more broadly—remain stuck in an annual renewal cycle, he notes
DeshCap uses the Basel Committee on Banking Supervision’s definition of a corporate customer’s operational risks across four main categories—people, processes, systems, and external events—and then into numerous subcategories that may vary across the organization’s businesses. It then analyzes how these risks are likely to be impacted by the company’s greatest risk exposures, whether natural disasters, cyber risk, or something else.
For a company with a hotel in a hurricane-prone location, for example, DeshCap may estimate a potential total loss of $50 million covering the entire risk spectrum, not just natural disasters.
“We then match up the operational data that comes from that analysis against the fine print of the company’s insurance policies,” says Barnachawy, “so we can bring down the overall quantum of operational risk in the business.”
AI Challenges
New technologies such as AI can help companies collect and analyze the data needed to identify and measure their risks more efficiently. Similarly, insurers and brokers are using AI to collate data and compare documents, such as a client’s asset registers in anticipation of annual renewals.
As corporate clients continue to move AI from the testing phase into their operations, however, insurers could face new challenges, Delchar warns. An insured company’s chatbot might hallucinate and give incorrect information to a customer, who may then act on it and choose to litigate.
“How insurable is that?” says Delchar. “AI is now in a relatively early stage of sophistication, so what are the checks and balances companies have in place to check what AI is doing?”
In the year ahead, insurers’ clients will likely find themselves interacting increasingly with AI agents, since the industry’s aging workforce is prompting insurers to explore how to augment humans to achieve greater scale and improve the customer experience. Moore foresees the highly regulated industry continuing to rely on humans to make key decisions, although regulators in each state of the US are mulling the issue and may arrive at different approaches.
Much of the data used in risk analysis is unstructured, Moore notes, ranging from financial statements to engineering documents to videos, and AI’s large language models can effectively convert that data into insights about potential risks. But it may be a while before the efficiencies promised by AI impact how insurers set rates. Ultimately, she says, insurers’ pricing models are based on actuarial concepts and constructs that rely on historical information.
“AI models can improve insights” about future risks, Moore says, “but the actual pricing models depend, as ever, on actuarial models that are backward looking.”
