1. Claims population

Heads of claims were asked to provide details of the seniority, roles and experience levels within their claims teams.

Initial signs suggest a shift in the demographic profile of the claims community; however, population data should be interpreted with caution, as the dataset does not include all managing agents.

The data indicates a notable increase in new entrants, coupled with a downward shift in the proportion of individuals with over seven and fifteen years’ experience, reinforcing anecdotal concerns among heads of claims regarding the loss of experience.

Survey respondents were asked to give details of how they attract and retain talent in claims.

The risks and opportunities identified by heads of claims over the next 3–5 years remain broadly consistent with those of 2023, but with a sharper and more specific articulation of where the pressure is greatest. Three themes have come into clearer focus: the impact of AI on skills and workforce planning; the potential loss of deep expertise as experienced professionals retire; and continued pressure on the mid-career talent segment.

When asked whether their view of future risks and opportunities had changed, heads of claims pointed to two overriding factors: the pace of technological change and increased automation, and a perceived reduction in knowledge and expertise within the market. Together, these are reshaping expectations of what good looks like in a claims professional.

The skills most frequently cited as critical for the future reflect this shift:

• data analytics
• technical expertise
• customer communication
• relationship management
• portfolio management
• adaptability and strategic thinking.

The presence of both technical and relational skills on this list is notable, suggesting that while technology will change how claims work is done, the human dimensions of the role remain highly valued. The challenge for the market is developing professionals who can credibly deliver both.

4. Risks and opportunities

The LMA International Bodily Injury Index (BII) provides a comprehensive view of bodily injury compensation outcomes across a broad international sample. By applying a consistent set of injury scenarios across multiple jurisdictions, it enables like-for-like comparison of indicative award values and helps track how those values change over time.

The Index is designed to be used by underwriters, claims professionals and risk specialists in managing agents to help in understanding bodily injury exposure across jurisdictions. It does not predict individual claim outcomes; actual claims depend on the specific facts of the case, liability, evidence, policy terms and local procedural requirements.

This overview focuses on interpretation and context. Detailed figures, jurisdiction-level data and scenario-specific analysis are available to subscribing managing agents.

The value of the International Bodily Injury Index lies in its ability to cut through anecdote and provide a consistent, evidence-based reference point across jurisdictions.

– David Fitzpatrick, Chair of the International Liability Business Panel

Purpose and value of the index

The International Bodily Injury Index provides a structured, transparent reference point for understanding how bodily injury awards vary internationally and how the sums awarded evolve over time. By applying consistent scenarios across a wide jurisdictional sample, the Index supports more informed discussion of claims cost trends and their drivers.

The 2024 to 2025 comparison establishes a baseline against which future movements can be assessed, allowing meaningful differentiation between gradual change and genuine structural shifts as the dataset develops over time.

Figure 1 note: Jurisdictions not shown may still be included in the Index but do not have complete data for both years yet.

What the data shows from 2024 to 2025

Data note

This analysis reflects the dataset as at 15 January 2026 and is based on jurisdictions with values available in both 2024 and 2025 at that time. The Index is updated periodically as additional jurisdictional data becomes available. Subsequent updates may expand the like-for-like sample and refine the distribution of movements.

Coverage for year-on-year analysis

Across the eight scenarios, between 59 to 60 jurisdictions currently have values available in both 2024 and 2025, depending on data availability for the specific scenario. These jurisdictions form the basis for year-on-year movement analysis.

Typical year-on-year movement

Across jurisdictions where values are available in both 2024 and 2025 (the like-for-like set used for year-on-year comparison), the median year-on-year change is modest:

• Minor or superficial injuries: increase of 1%.
• Broken limb injuries: increase of 1%.
• Serious and catastrophic injuries (including amputation, loss of sight, psychiatric injury, brain injury and paralysis): increases typically between 2% and 3%.
• Fatal outcome scenario (full awareness of death): increase of 3%.

The distribution shows that most jurisdictions experienced incremental, rather than dramatic movement, between 2024 and 2025. However, a small number of jurisdictions show more pronounced year-on-year movements. Ten to fifteen jurisdictions fall into the “large increase” band (more than 10% year-on-year), depending on the scenario.

For example, in the minor or superficial injuries scenario, jurisdictions such as Uzbekistan and Japan show large year-on-year increases, while in the paralysis scenario, large movements are seen in jurisdictions such as the UAE and Jamaica. These larger shifts are typically linked to discreet local developments, such as legislative change, court practice shifts or updated valuation reference points.

Distribution and variability

Although median movements are low, the distribution of outcomes is wider for some scenarios. Interquartile ranges vary materially, particularly for minor injuries and paralysis. This reflects differing local dynamics rather than a uniform global trend.

Across scenarios, a significant proportion of jurisdictions show no year-on-year change at all. Depending on the scenario, between 15 and 25 jurisdictions remain unchanged within the like-for-like sample.

Direction of change

Increases are more common than decreases, but decreases do occur:

• In minor injury scenarios, 53% of jurisdictions showed increases, 5% decreased and 42% were unchanged.
• In fatal outcomes scenarios, 64% increased, 5% decreased and 31% were unchanged.

This demonstrates that bodily injury awards are not moving uniformly in a single direction across all jurisdictions.

The charts below are derived from the International Bodily Injury Index. They illustrate high-level patterns and do not display detailed figures or jurisdiction-level values. The two charts present complementary views of year-on-year movement. Figure 1 shows the scale of change by grouping jurisdictions into movement bands, while Figure 2 shows the overall direction of movement across jurisdictions.

What the data shows is not a single international story, but a pattern of stability for many jurisdictions alongside targeted areas of change driven by local legal and economic factors.

– Chris Mather, Senior Executive, Technical Underwriting at the LMA and Secretary of the International Liability Business Panel

Step changes and outliers

As set out above while the median year-on-year changes are modest and most jurisdictions fall within the stable or moderate movement bands, a smaller group falls into the “large increase” band, defined as movements of more than 10% year-on-year.

Examples observed in the data include:

• In a small number of jurisdictions, year-on-year movements fall into the “large increase” band (more than 10% year-on-year), representing a minority of the like-for-like set. These more pronounced shifts tend to be concentrated in higher-severity scenarios and are more consistent with discrete local developments than with gradual inflation.
• Larger step changes are more commonly seen in higher-severity scenarios (for example, paralysis and amputation) where local valuation frameworks can shift materially.
• Decreases are less common than increases and are limited to a handful of jurisdictions per scenario, with some of the more notable decreases appearing in minor or superficial injury scenarios.

These outliers can skew the overall average when year-on-year percentage changes are summarised across jurisdictions. The analysis compares two figures per jurisdiction (2024 and 2025) and calculates a year-on-year percentage change for each jurisdiction in the like-for-like set. These jurisdiction-level percentage changes are then summarised across that set. To avoid distortion from outliers, this overview presents the median and the distribution
of movements rather than relying on a single average figure. Larger shifts may also point to discrete local developments, such as legislative change, shifts in court practice, updated damages guidelines and valuation reference points, influencing outcomes in particular jurisdictions.

Why this is significant

Limitations and context

The scale of the Index is a strength, but it requires careful interpretation:

• Blank fields reflect unknown data, not a 0 value.
• Figures are indicative benchmark estimates, not claim predictions, developed through specialist legal input and jurisdictional insight.
• Local context remains essential and is captured through jurisdiction-specific notes where available.
• The Index is not a pricing or reserving tool and should not be used as a substitute for claim-specific analysis.
• The Index is not intended to set market benchmarks, tariffs or pricing levels, nor to replace jurisdiction-specific legal advice.

Key definitions

Bodily injury scenario

A standardised, injury outcome used consistently across all jurisdictions. Each scenario is based on the same claimant profile and injury description to support comparability.

Like-for-like comparison

A comparison between years that includes only jurisdictions where a value is available for the same scenario in both 2024 and 2025. Jurisdictions without data in one of the years are excluded from year-on-year analysis for that scenario.

Year-on-year (YoY) change

The percentage change between the 2024 and 2025 indicative values for the same scenario and jurisdiction. Year-on-year analysis is calculated only on a like-for-like basis.

Local currency

Indicative values are shown in the currency used within each jurisdiction. Figures are not converted into a single reference currency to avoid masking local economic and legal dynamics.

Blank data field

A blank field indicates that the relevant value is unknown or unavailable for that jurisdiction and year. It does not mean that the value is zero or that no award is payable.

Indicative value

A benchmark figure reflecting informed market understanding for a defined scenario. Indicative values are not claim outcomes and should not be interpreted as pricing, tariffs or guaranteed awards.

Jurisdiction

A country or legally distinct legal system included within the Index. In some cases, (for example, Australia and Canada, we have dealt with different states within the country as their rules and awards can differ materially.)

Methodology (how year-on-year movement is calculated)

Year-on-year movement is calculated using the LMA International Bodily Injury Index’s standardised scenarios, comparing 2024 and 2025 values on a like-like basis. For each scenario, jurisdictions are included only where values are available in both years.

The year-on-year percentage change is calculated as:

In 2025, we rang the bell on a critical challenge facing the London market: the female underwriting leadership pipeline was going backwards. This year, we move the conversation forward.

The 2026 Underwriting Talent Summit will focus on progression. If we are building a pipeline of talented female underwriters, how do we ensure they reach the most senior leadership roles? What barriers remain? What do future female leaders need from their employers? And what can senior underwriting leaders do differently to create lasting change?

Bringing together CEOs, CUOs and senior underwriters from across the market, this summit will explore the practical actions, leadership behaviours and organisational cultures that enable talent to thrive and progress. Our panellists and speakers will reflect on the reality of building a career in the London market: the achievements, the challenges and how they created their own pathway to leadership.

Speakers

Further speakers will be announced over the coming weeks.

Executive summary

This paper examines the use of indexation in the review of portfolio trackers and documents methods to measure it. Currently, there is no universally accepted market standard for measuring indexation. As heightened attention has been placed on the performance of portfolio trackers, it is becoming increasingly important to accurately report and be able to compare indexation across facilities so that targeted action can be focused on areas that require it most. It can also form part of contractual arrangements between a portfolio tracking coverholder and underwriters, enhancing the need for calculations to be accurate and transparent. The following discussion suggests a framework for measuring indexation that could be adopted by underwriters and coverholders if thought appropriate.

What is indexation?

Portfolio trackers are supposed to ‘track’ or ‘index’ an entire portfolio of risks, sometimes across many classes of business. Indexation is a measure used to track how well capacity is being deployed across a portfolio. Indexation can also be used to detect anti-selection, i.e. where capacity is more likely to be deployed behind hard-to-place/lower performing risks as opposed to being deployed across all in scope risks. It is not relevant for coverholders operating other follow facilities in which the aim is something other than to track an identified portfolio of risks.

Indexation is typically reported as a percentage – premium or policy count in respect of bound business divided by that which is in-scope for the portfolio  – where perfect indexation is 100% (see below). Other terms commonly used in relation to indexation include ‘utilisation’ and ‘adoption’. While these may sometimes refer to distinct metrics, they are frequently used interchangeably, highlighting the need for clarity, so all parties understand what is being measured and reported.

It may be the coverholder’s responsibility, if set out in the agreement, to gather all the relevant information and data to calculate indexation in an appropriate manner. The accuracy and completeness of this should be an area of audit or due diligence of the underwriter.

Why is it important for portfolio trackers?

Portfolio trackers should, in theory, balance the good with the bad as they should reflect the in-scope portfolio. If capacity is unfavourably deployed towards historically poor performing risks rather than evenly across the portfolio, then the loss ratio of the portfolio tracker may be impacted, affecting the long-term sustainability of the facility. Conversely, portfolio trackers should not tend to be selective toward the best-performing risks, as they aim to track the entirety of a portfolio that supports the needs of brokers to offer capacity to their clients.

Portfolio trackers succeed through diversification of risk, often by class of business and/or geography. The initial assessment of a portfolio tracker opportunity by carriers involves reviewing the composition of an existing portfolio and future development projections. Any large deviations from the anticipated portfolio mix can impact the success and therefore long-term sustainability of the facility. For example, a facility may have been presented as being 65% short tail, 35% long tail, so if indexation in long tail casualty lines is poor, then the carriers’ appetite may no longer be satisfied by the facility.

Furthermore, large deviations from the anticipated portfolio mix will result in inaccurate actuarial pricing: portfolio trackers typically apply a development pattern based on the projected business composition and derive an expected ultimate loss ratio accordingly. Significant deviation from the anticipated portfolio, i.e. indexation at significantly less than 100%, can lead to an inappropriate ‘loss pick’, reducing a carrier’s ability to accurately reserve.

Monitoring how well indexed a portfolio tracker is relies upon coverholders regularly reporting on indexation metrics, so that any areas of concern around low levels of indexation can be identified early, and a proactive approach can be taken to address the concern.

Methods of measuring indexation

As mentioned above, a standard formula for indexation is set out below: a variety of different metrics can be calculated by replacing the X in the formula (e.g. most likely to be premium or number of risks in the potential portfolio).

The fraction denominator within this formula (i.e. the bottom number of the fraction) is the ‘in-scope’ portfolio – getting this right is difficult and critical – and the top line is the portfolio that has been bound.

The process of measuring indexation can be broken down into three main elements: scoping methodology; reporting metrics; and reporting of missed opportunities.

Scoping methodology

Successful indexation measurement relies on the accurate and transparent assessment of in-scoping risks. Underwriters and coverholders have varying approaches as to what is considered in-scope for indexation. Illustrative examples of “outcome reasons” and categorisation for in/out of scope are given below.

In scope (included within the denominator)	Out of scope (excluded from the denominator)
Not taken up due to broker choice	100% single markets
Not offered due to late submission	Unapproved leaders/risk codes
Not offered due to delayed referral response	Policy periods exceed maximum for facility
Not offered due to limited aggregate capacity	Excluded territories/insureds/failed ESG
Client declinature – pricing	Suitable alternative option taken up
Client declinature – insurer continuity/loyalty	Broker lost the account at renewal
Client declinature – outstanding claims	Referral declined
	Limits in excess of facility maximum $/% line size

On the above table, where the parameters of the facility have restricted the placement, this should be deemed ‘out-of-scope’ as there is nothing the coverholder can do to place the business in the facility. ‘In scope’ reasons are those over which the coverholder has some degree of control or influence over the placement or otherwise of the business in the facility.

It may be that it is unclear as to whether a given item should be deemed in or out-of-scope for indexation purposes, e.g. reduced lines/missed opportunities due to limited aggregate capacity. The scoping exercise is typically carried out jointly between the Lead Underwriter and coverholder, and the in-scoping criteria shared with the panel participants for transparency. Following this in-scoping exercise, the quality and range of reporting metrics then add quantifiable results that can be monitored.

The in-scoping process is most effective when it occurs both prior to the facility period and during, by class of business. An initial assessment of the in-scope portfolio is necessary for parties to consider its likely performance. Underwriters should conduct an ‘ex-post’ in-scoping at regular intervals once the true subject portfolio is known with certainty. This allows for unknowns, such as new or lost business or any uncertainty or inaccuracies in the in-scoping process, to be corrected prior to the calculation of the indexation. This is necessary to avoid corrupting indexation measures with irrelevant data.   

Reporting metrics

Some coverholders measure indexation by combining multiple metrics to aid high-level tracking over time. However, it should be remembered that there is no single measure that can fully convey the indexation of a portfolio. True indexation requires assessment using a variety of different metrics, as the success of one metric does not necessarily result in good indexation throughout. Examples of metrics that can be used to measure indexation include the following:

• % Client Indexation – proportion of in–scope clients that bound at least one policy.
• % Policy Indexation – proportion of total in–scope policies (UMRs) that were bound.
• % Premium Indexation – proportion of in–scope premium that was bound. This should ideally be measured using GWP for the portfolio tracker’s participation.
• % Layer Indexation – proportion of in–scope layers that were bound. This is typically done on a per-programme basis and then averaged across a portfolio.
• % Signings – proportion of offered capacity that was signed (written vs. signed line).
• Average % of Order (a) – the average signed % line relative to the total order of the policy. Note: relevant for portfolio trackers that offer lines based on order size.
• Average Signed % line (b) – the average signed % line. Note: relevant for portfolio trackers that offer lines based on whole only.
• Average Limit utilization – the average $ limit deployed across a programme.

The final three metrics listed above monitor the capacity deployment and can be used to assess any areas where capacity is being unduly restricted by the line size or capacity limits (e.g. where the percentage line is restricting the $ use or vice versa).

Policy and premium indexation measures are the most typical measures in use today. The more granular and regular the reporting, the easier it is to identify trends to proactively manage any actions required to improve indexation.  

Reporting of missed opportunities

The final element of indexation that may be reported on is in-scope opportunities that were ultimately not taken up. For example, if a portfolio achieves 98% indexation, reviewing the reasons for the remaining 2% through an ‘NTU’ (not taken up) assessment provides context as to why they were not taken up, so these can be addressed to improve indexation.  

Although out-of-scope missed opportunities are not directly relevant to indexation monitoring, tracking metrics will increase transparency and identify areas where expanding the facility’s scope may address gaps, such as risks involving unapproved leaders or risk codes.  Underwriters and coverholders could also assess the relative performance of the out-of-scope business. 

Conclusion

Reporting indexation with multiple metrics by class of business and in total provides a comprehensive picture. Regular reporting allows underwriters to fully assess the performance of the capacity deployment. Coverholders should have the ability to gather all the necessary information and data to calculate indexation in an appropriate manner.

Appendix A illustrates an example indexation report for a specific month across various metrics. The first two columns demonstrate the difference between the initial in-scope EPI prediction for the month relative to the revised EPI following any adjustments to the in-scoping exercise. At a high level this comparison allows recalibration of the indexation figures against the revised EPI, while still reporting on the initial EPI for the month so that carriers are aware of any movements against original projections.  

Appendix B then shows an alternative view where a single indexation metric is selected, with variability shown by month per class of business. This allows for visibility of any trends in indexation over time, highlighting any areas of concern in need of review. 

Appendix C goes into greater depth around the challenges associated with calculating indexation.

Appendix A – Indexation reporting for each month independently

An illustrative table showing the individual metrics by class of business, with the ability to select a timescale (e.g. monthly, quarterly, annually, year to date). Regular reporting will add context to the indexation of the portfolio by assessing multiple indexation metrics simultaneously.

Select reporting period ▽
	In-scope premium (initial EPI)	Revised In-scope premium	% Client indexation	% Policy indexation	% Premium indexation	% Layer indexation	% Signings	Average % Order/Average signed % line	Average Limit utilisation
International Property
North American Property
Construction
Product Recall
Power
Marine Cargo
Marine Hull
Marine Liabilities
Specie
Energy
International Casualty
US Casualty
Healthcare and Other Casualty
Professions
Commercial D&O
Financial Institutions
Other Financial Services
Cyber
…

In-scope premium (Initial EPI): Based on the initial in-scope analysis, what the estimated premium income (EPI) was

Revised In-scope premium (ultimate EPI): With hindsight, what ultimately ended up being in-scope:

• Following the removal of any risks that were no longer in-scope (e.g. policies that moved to 100% single markets)
• Following the addition of new business that fits the in-scoping criteria

Appendix B – Indexation reporting demonstrating trends by indexation measure

A second illustrative table showing the indexation measurements over time (per month independently), with a drop-down to select the indexation measure you wish to review, e.g. client indexation, policy indexation, premium indexation etc. This report will give visibility over how indexation is tracking with time per class of business.

Appendix C – Example process and challenges

Example process flow

1. The process of determining the in-scope portfolio starts with the pipeline of upcoming renewals (and to the extent possible, new business) and assessing their suitability for the tracker against the underwriting parameters. 
2. All risks are appropriately labelled based on the outcome, so that any risks that did not bind have reasons as to why they were not taken up.
3. These NTU reasons are categorised as either in-scope or out-of-scope. Some risks that previously met the in-scope criteria may not have been eligible for renewal due to changes in the underlying factors, e.g. change in leader that was no longer in scope. Additionally, any business lost by the broker needs to be de-scoped from this exercise.
4. Retrospectively review any additional risks that were bound that may not have been in the initial in-scope pipeline, including new business won by the broker, or newly in-scope risks that previously did not meet the underwriting parameters.
5. A full in-scoping review from pipelining stage to post-bind reconciliation is required to fully assess the in-scope criteria against which indexation needs to be measured. This requires there to be consistent labelling of outcome reasons. Parties need to pay attention that confidentiality and competition concerns are not engaged. 

In a facility’s first year, there is often insufficient data to form a perfect in-scoping exercise, making the ex-post monthly review of the in-scoping vital in understanding true indexation as opposed to inaccurate in-scoping. Additional controls/sign-offs may be required during this period to prevent significant deviations from expected early on. 

Should 100% indexation be achievable? 

There will always be optionality for clients as to whether or not to accept capacity from a portfolio tracker, meaning that indexation will never reach 100%. As a result, many trackers now offer incentives to clients, such as discounts on the lead premiums, encouraging the use of the tracker capacity. The end client cannot be forced to purchase cover from tracker portfolios but can be encouraged and incentivised to do so, subject to any duties that the broker will owe to the insured. The coverholder can also take steps such as communicating the benefits of utilising the tracker to both brokers and clients in areas where indexation is poor.  Stating the obvious, giving discounts to lead premiums and additionally payments for participating on the tracker will mean that the portfolio will track at a higher loss ratio than an open market lead on the same business.

Should client declinatures count negatively against indexation? 

As discussed above, indexation is a measure of how well the coverholder is deploying the capacity. If a client chooses to decline the portfolio tracker’s capacity, should the coverholder be penalised for this, given it is out of the coverholder’s control? Some suggest that client declinatures should be de-scoped and not be included within the denominator. This would support 100% indexation being achievable to the coverholder based on factors that are within their control.

However, this may devalue the indexation metric. Furthermore, clients may be sceptical about the optionality of portfolio trackers should they see that portfolio trackers are achieving 100% indexation, as clients should always have the final decision as to whether they opt to take it up. However, a proposed workaround for this is having separate indexation measures including and excluding client declinatures, so that declinatures outside the control of the coverholder are not taken into account. Overall indexation may be 85%, but excluding client declinatures, coverholder indexation may be much higher, again adding further context to the final metrics.

Should reduced lines following aggregate limit restrictions count negatively against indexation?

Contractual aggregate exposed limits and modelled limits are often set upfront based on the anticipated portfolio, so that underwriters can manage their exposure and purchase appropriate reinsurance. Coverholders monitor utilisation relative to these limits, and if they are nearing these limits, they may cut back on line sizes or decline risks in some regions. If the coverholder was not proactive in assessing changes in the portfolio throughout the year and addressing any requests to increase limits in advance of any concerns, then any remedial action of reducing line sizes/declining risks may count negatively on indexation.  

If, however, a small limit was deliberately set by underwriters due to appetite, then any reduction in capacity as this limit is neared may not count against the coverholder from an indexation perspective (though the line size should have been reduced for that class/region to manage this in a controlled way).  

How should premium indexation be calculated? 

Ideally, premium indexation should be measured using the GWP for the portfolio trackers’ participation. For large programme limits, the % line sizes will often be restricted by the $ limits, and the in-scope premium for the portfolio tracker will depend on the programme structure. What is deemed the in-scope GWP to the portfolio tracker can therefore change significantly from the pipeline stage until the risk(s) have bound, as policies become in/out of scope and premiums are negotiated. For NTUd risks, the only way to determine this in-scope GWP would be to quote the portfolio tracker lines retrospectively based on the bound policies, which would often be a manual exercise and difficult to regularly review.  

Where premium indexation is presented using 100% GWP for the in-scope policies, rather than calculating the portfolio trackers’ participation based on realistic %/$ lines, this does not truly reflect the premium indexation across a portfolio as the outcome would result in very high premium programmes unfairly skewing the premium indexation results even if the tracker line sizes in scope were small. 

Acknowledgements

The LMA would like to thank the members of the CUO Committee and the CUOC Enhanced Underwriting Subcommittee for their contribution to this report. Thank you in particular to Colum D’Auria and Elinor Taylor.

The LMA has contributed an Expert Analysis chapter to the International Comparative Legal Guide (ICLG) to Insurance & Reinsurance 2026.

The chapter, Navigating the risks of enhanced underwriting, examines the risks associated with enhanced underwriting and is now available to read below.

Enhanced underwriting models are reshaping how risks are selected, priced and bound, offering significant operational efficiencies. To support these varying models, managing agents continue to rely on established placement frameworks as the underlying contractual infrastruc­ture. This chapter explores how these long-standing contractual mechanisms interact with new system-driven underwriting processes and where key risks may emerge.

If you have any questions, please contact me.

In 2025, the LMA worked closely with managing agents, Lloyd’s, regulators and the wider (re)insurance ecosystem to enable the success of our members and the market. We advanced the simplification and digitisation of the market, strengthened our policy and regulatory influence, supported and represented Lloyd’s underwriters, and continued to invest in the talent pipeline that will shape the future of the Lloyd’s market.

Here are the highlights.

Introduction

Artificial intelligence (AI) systems and tools have been an extremely hot topic in insurance for several years and are likely to remain so for the foreseeable future. Many boards of Lloyd’s managing agents have recently cited AI risk as second only to geopolitical risks on risk registers.[1] In order to explore some of the key themes emerging from insured AI risk, the LMA conducted a survey of members, asking for views across a range of potential loss scenarios.

Primarily, the survey was intended to research:

• LMA members’ views on whether insureds were using AI, right now, in ways that could cause or contribute to claims
• which loss scenarios LMA members were most concerned about
• LMA members’ views on whether insureds were managing the risks adequately
• members’ views on the potential scale of losses.

Definition of AI

The definition of AI used in the survey describes an AI system as “software that is automated and generates information outputs based on a statistical model of data taken from similar or informative scenarios.”

In the survey, respondents were encouraged to think about AI very broadly, including any computer system able to perform tasks that normally require human intelligence, under varying and unpredictable circumstances, without significant human oversight, such as visual perception, speech recognition, decision making and language translation. AI systems could include large language models, machine learning software and deep learning software.

Use of cyber affirmations and exclusions in policy wordings

The survey also collected some general views on whether AI exposures might typically be covered or excluded in policy wordings. The LMA previously explored this issue and briefed the LMA Chief Underwriting Officers’ Committee in 2024 on a high-level analysis of LMA model wordings exposure to a list of AI loss scenarios. An updated summary table of this work is provided in Annex 2.

Some high-level questions on the use of exclusions were included in the survey but are not reported in this summary. On review, it was not appropriate to report on generalised observations about coverage/exclusion of losses without narrowing down the scenario(s) to an exact set of circumstances and the exact clause that was used; this is outside the scope of the survey and would be more accurately researched via case studies.

Disclaimers

The observations reported in this summary do not relate to specific losses, wordings or exclusions. An authoritative view on coverage or exclusion of losses arising from a specific loss would require a review of the facts and circumstances and application of the policy wording(s) used.

Also, we did not explore the protective effects of AI in this survey, taking into account the benefits that could accrue to insureds and (re)insurers arising from the use of AI, which could be material.

---

[1] Lloyd’s, 2025. Emerging Risks Report.

Methodology

A 31-question survey was developed and issued by the LMA to members in April 2025. Members were invited to select any of the listed scenarios and complete the question set and were encouraged to submit responses for any of the scenarios within their area(s) of expertise.

The LMA developed a lengthy suite of AI loss scenarios: 65 in total, across 10 high-level classes of business. Some scenarios were invented for this exercise and some were based on events that have (or may have) already occurred. The scenarios were validated with the LMA’s underwriting committees, to ensure that each scenario was believed to be technically viable, even where the probability of a particular scenario occurring was suspected to be very low.

The scenarios were deliberately broadly drafted, to try to encompass a wide range of AI use cases/loss scenarios and attract as many responses as possible, on the basis that this survey is exploring AI risks at a high-level only. For example, in the Professional Indemnity Scenario 3.1 “AI produces erroneous advice/service to clients, causing a loss”, the type of advice/service provided is not specified as the intention was to capture any type of advice or service provided to any type of professional client. This approach does have its limitations in terms of the degree to which the views and opinions captured might be generally applicable, but the intention was to capture high-level views on generic AI loss scenarios for major insurance products.

Respondents

144 responses to the survey were collected in Q2 and Q3 2025, 94% of which were from underwriters. The distribution of responses by job role was:

• Leadership/C-suite: 12%
• Head of Class/Senior Underwriter/Senior Manager/Senior Product Manager: 33%
• Manager: 31%
• Team member: 25%

Many respondents advised that the survey questions were difficult to answer. Not enough responses were collected to analyse every scenario included in the survey. Therefore, this summary focuses on the four scenarios with the highest response rate, with a minimum threshold of six responses per scenario, though up to 19 responses were received for some scenarios.

Respondents reported the survey as difficult to complete, for two main reasons:

• Insurers are remote to insureds’ use of AI; it is not always clear whether an insured is using AI in the ordinary course of their business and little data on AI usage is currently collected during underwriting, so survey questions exploring this usage have relied on underwriters’ views and opinions rather than data.
• There is a lack of claims data on AI contribution to losses, again meaning that the survey responses are informed by views and opinions rather than data in most cases.

Future research may be able to draw more directly on underwriting and claims data. However, the key observations arising from this survey still address the LMA’s overall goal of contributing to the sum of knowledge on AI risks and sharing the insights arising from this research with LMA members.

Scenario viability, magnitude and overall impact ratings

Each of the four scenarios that feature in this analysis have been given a viability rating and a magnitude rating, to summarise respondents’ views on these core issues. The ratings are based on an assessment of respondents’ views on key questions in the survey, regarding whether the scenarios are believed to be technically possible or not, and the level of concern about the loss potential. An overall impact rating for each scenario has also been produced by multiplying the viability and magnitude ratings together.

The viability rating is an assessment of whether the scenario is believed to be technically possible or not, based on responses to questions that probed:

• whether AI was thought to be in use in live situations, in the way set out in the scenario
• if so, whether it was widely or narrowly deployed
• whether AI was making “real-world” decisions.

The viability rating is expressed on a five-point Likert scale, using weighted average responses:

	Viability rating	Rationale
1	Highly unlikely	Responses indicated that this scenario was highly unlikely to occur
2	Unlikely	Responses indicated that this scenario was unlikely to occur
3	Plausible	Responses indicated that this scenario was plausible
4	Likely	Responses indicated that this scenario was likely to occur
5	Highly likely	Responses indicated that this scenario was highly likely to occur

The viability rating also touches on potential frequency; the higher the score, the more likely respondents felt the scenario was to occur and produce a real loss event.

The magnitude rating summarises respondents’ opinions on the potential scale of loss, should the scenario occur for real. The rating is drawn from opinions received on the question exploring the potential scale of loss, expressed as a percentage of typical limits of cover. The magnitude rating is expressed on a five-point Likert scale, using weighted average responses:

	Magnitude rating	Rationale
1	Low value	Responses indicated that this scenario could produce low-value losses only
2	Minor	Responses indicated that this scenario could produce minor losses
3	Medium	Responses indicated that this scenario could produce medium losses
4	Sizeable	Responses indicated that this scenario could produce sizeable losses
5	Full limit	Responses indicated that this scenario could produce full-limit losses

An overall impact rating for the four loss scenarios has been produced by multiplying the viability rating and magnitude rating of each scenario together. This rating is intended to summarise the respondents’ opinions on the potential overall impact of a loss scenario, reflecting views on both the viability of the scenario and potential scale of loss in the event the scenario occurred.

	Overall impact rating	Rationale
<5	Minimal	Overall impact (per loss) estimated as potentially minimal
<9	Low	Overall impact (per loss) estimated as potentially low
<13	Moderate	Overall impact (per loss) estimated as potentially moderate
<17	Major	Overall impact (per loss) estimated as potentially major
17+	Severe	Overall impact (per loss) estimated as potentially severe

Note: The ratings are intended to be indicative of the opinions of respondents to the survey; they are not based on actuarial calculations or driven by claims, exposure or risk management data.

Key loss scenarios

Detailed analysis and commentary is provided below on four scenarios:

• Professional Indemnity scenario 3.1: “AI produces erroneous advice/service to clients, causing a loss.”
• Product Recall scenario 4.13: “Product Recall due to property damage and/or bodily injury caused by defective products designed and/or manufactured by AI, e.g. food contamination.”
• Accident and Health (A&H) scenario 1.1: “self-driving car error causes injury to passenger: first-party A&H policy could respond in first instance.”
• Cyber scenario 10.1: “System downtime caused by AI malfunction and resulting business interruption.”

Responses to each of the four scenarios have been collated into three areas of analysis:

1. The use of AI by insureds and growth potential in the near future.
2. Risk management.
3. Potential impact on losses.

For each of the four scenarios, a rating for viability, magnitude and overall impact has been provided. Please see Section 3 above for further information on the methodology of these calculations, as well as commentary on the trends and data points of interest.

Treatment of AI risk in model wordings

In 2024, the LMA undertook a review of the potential AI coverage position in the LMA’s suite of model cyber clauses. A summary table is provided in Annex 2, updated to include model cyber wordings published up to November 2025.

In this review, AI is recognised as a sub-set of software and therefore falling within the definition of Computer System, widely used in LMA model Cyber clauses:

Computer System means any computer, hardware, software, communications system, electronic device (including, but not limited to, smart phone, laptop, tablet, wearable device), server, cloud or microcontroller including any similar system or any configuration of the aforementioned and including any associated input, output, data storage device, networking equipment or back up facility, owned or operated by the Insured or any other party.

The potential AI coverage position, for each model clause, is indicated with a “low”, “medium” or “high” rating, based on assessment of whether the clause is:

• Low: fully excluding cyber risks.
• Medium: an exclusion of cyber risks with a write-back of cyber coverage or an affirmation of cyber risks coverage with a limitation.
• High: a full grant of cyber coverage with no cyber-specific restrictions but subject to the other terms and conditions of the policy.

Disclaimer: The summary table in Annex 2 is intended to be indicative of potential AI (cyber) risk; the actual risks present in any contract will depend on the wording(s) used and the circumstances. The LMA accepts no liability to any party in respect of the information provided herein.

Annex

Annex 1: Full list of AI loss scenarios included in 2025 survey

Annex 2: List of LMA cyber wordings categorised for AI exposure

The LMA and Lloyd’s Underwriting Talent Summit brought together 180 senior underwriting leaders to shine a spotlight on the trends impacting the pipeline of female underwriting leaders.

The event’s purpose was threefold: to share the lived experiences of female underwriting leaders, to identify the barriers slowing progression and to outline the clear steps that we can take forward as a market.

Below, we outline the individual actions that CEOs, CUOs and underwriting leaders can take to drive meaningful change to build the female underwriting leadership pipeline across our market.