IFoA GIRO Conference 2026

Many UK residents living in high flood risk areas have enjoyed the benefit and peace of mind of affordable flood insurance for many years, which has been kept affordable through the current Flood Re scheme. This will end by 2039, UNLESS… 

With ever more homes and businesses becoming exposed to potentially unaffordable premiums as climate change progresses, and as the current scheme (which only covers primary homes built before 2009) draws to a close in just over a decade, the UK needs to develop a new solution. 

It’s unclear what, if anything, anyone else is doing about this. Given our skillset and technical expertise, should actuaries take the lead? 

Our working party, supported by the IFoA Policy team, will present our latest thinking, research and discussions, including engagement with Government and other stakeholders to promote early development of a broader and more robust, appropriately funded solution.  

We will share our design for a potential post-2039 solution, looking at potential funding, governance and coverage of the growing flood protection gaps. We will set out possible ways of linking flood defence management and funding with the new insurance solution. 

We also aim to share the results of our latest planned survey, which will focus on how insurers are utilising CAT modelling tools to allow for ever-changing flood patterns. It will compare and contrast how insurers make use of CAT models in their flood pricing and underwriting decision-making processes, compared with other perils. 
 
Assumed knowledge level: 
No specialist knowledge required 

Speakers: 

Richard Stock, Flood Working Party
David McNamara, Flood Working Party
Rachel McCallion, Flood Working Party

The global P&C market has entered a phase of rapid and intense softening. Is this just another turn in the cycle, or is something truly different? 

Presented by the IFoA GI Pricing Research Group ('PRG'), this session will provide its latest view on where we really are in the underwriting cycle, building on well-received GIRO presentations in 2024 and 2025. 

1. Market overview 2026 

We will present recent and projected market performance statistics, analysing the main movements in pricing and profitability, and focusing on the forces that shape those outcomes — inflation dynamics, capital flows, competitive behaviour, and secular shifts.   

We will look at this in two main segments: 

• UK retail business (motor, home, and commercial lines) 
• Lloyd's/London market/global P&C and reinsurance

2. Issues register and horizon scan 

Alongside the state of the market review, we will present our annual issues register and horizon scan analysis. This will highlight the challenges and opportunities for pricing teams, now and into the future, shaped by emerging technology developments, the changing geopolitical and socioeconomic context, and regulatory and conduct pressures.   

We will also showcase where active research across the IFoA and the wider actuarial community is helping to address these concerns. 

Assumed knowledge level: 

Prior knowledge required (ensuring a delegate working in that area learns something new)

Speakers:

Edward Plowman, GI Pricing Research Group
Mohammad Khan, GI Pricing Research Group
Andy Goldby, GI Pricing Research Group

AI data centres (AI DCs) are a new class of risk for insurers. Due to their design, power intensity and operational dependencies, they have a distinct risk profile that is fundamentally different from conventional data centres. 
 
The scale of the expansion of AI DCs represents a material source of risk, with large, complex construction projects and significant concentrations of value at single sites. Once the pace of development and network of indirect exposures are considered, these assets have the potential to become a meaningful source of accumulation risk.  

Their dependence on electricity supply, availability of the latest chips, enormous and interconnected financing structures and multiple counterparties creates a web of exposures spanning property, liability, cyber, professional indemnity and credit lines.  

The promise of tremendous premium volumes has AI DCs on every insurer’s business plans for the next 5 to 10 years, but insurers should consider whether existing underwriting approaches and risk frameworks are sufficient to capture this complex and evolving risk landscape. 

Summary: 

• What are the key components of AI infrastructure, their characteristics and most relevant casualty insurance covers? 
• How do exposures and risks grow and change through the lifecycle of a given AI DC? 
• What is the scope and scale of the AI infrastructure sector today, and what do forecasts suggest its future to be? 
• Where might large single- and aggregating- losses arise within the AI Infrastructure sector today and into the future? 
• How do we best view AI infrastructure risk from a climate liability lens? 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Michelle Gascoin, Liability Exposure Management Working Party
Matthew Lewis, Liability Exposure Management Working Party

The PRA ran the first dynamic general insurance stress test in May. This panel discussion between the PRA, Lloyd’s and selected insurers will reflect on the outcomes of DyGist through an actuarial lens. 
 
After a short introduction on the purpose and a summary of the exercise, the panel will focus on the key insights and implications of a dynamic stress, in particular how it challenges traditional actuarial approaches to capital, reserving and business planning. 
 
Discussion topics will include: 

• What were the objectives of DyGist, and have they been achieved? 
• When and how actuarial and risk teams were brought into the exercise, and how that worked in practice 
• How Lloyd’s dealt with specific features, such as the member structure 
• What we have learnt, including implications for capital setting post-event, how firms and Lloyd’s respond to new business opportunities, and operational challenges (for example, managing a surge in business plans) 
• What this means for how actuarial frameworks/processes may need to evolve to deal with dynamic, large/multi stress events. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Mirjam Spies, Lloyd's
Cian O'Brien, Lloyd's

The chain-ladder method has dominated claims reserving for decades, yet individual claims reserving, which exploits claim-level features to improve estimates, has not become established in practice. We attribute this to the absence of a methodology that is simultaneously practical, robust, and naturally connected to existing actuarial workflows. 
 
We present a new approach that restructures the chain-ladder algorithm to enable individual claims reserving. The key insight is reformulating factor estimation into a recursive one-shot projection-to-ultimate framework. Rather than rolling forward cumulative payments one period at a time, we estimate grossing-up factors that project each claim's latest observation directly to ultimate.  

This representation is mathematically identical to standard chain ladder on aggregate data, but creates a natural pathway for incorporating claim-level features through regression modelling. 
 
The approach handles arbitrary dynamic covariates, including claim status, claims incurred and case reserves, without requiring their extrapolation, removing the principal barrier in existing micro-reserving methods. It separates RBNS and IBNR claims through consistent cohort construction, providing a novel decomposition of total chain-ladder reserves.  

We present results on two real datasets benchmarked against known lower triangles. 
 
This session is aimed at reserving actuaries at all levels. No deep learning expertise is assumed; we show that linear regression can be surprisingly effective for micro-reserving. 

Assumed knowledge: 

Prior knowledge required (ensuring a delegate working in that area learns something new) 

Speaker:

Ron Richman, insureAI

The working party was set up to address the practical issues that came to light with the implementation of Solvency II in 2016. It was focused on carrying out a practical review of the Solvency II regime to help guide modifications to Solvency II post Brexit. 
 
The UK has now moved from Solvency II to Solvency UK. Our renewed scope is to understand the impact of this on the non-life sector. The working party will provide an update of the impact of Solvency UK two years post implementation, supported by a market-wide survey.  

We will also cover similar regimes being implemented globally, such as the risk-based capital, wider holistic impact of this change and what to expect in the longer term. 

Assumed knowledge level: 

No specialist knowledge required 

Speaker:

Amrita Pattni, Solvency UK Working Party

The growing adoption of credit insurance products, driven in part by banks increasingly transferring credit risk to the insurance world, raises a critical question: do actuaries possess the tools, frameworks and expertise to truly understand and manage these risks?  
 
Credit business, from single obligor exposures to structured credit, introduces a nuanced complexity. Data is often limited, expert judgement is heavily relied on, sensitivity to macroeconomic conditions is heightened, and dependency effects are more pronounced.  

Moreover, there is a risk of disconnect between deal-level underwriting and portfolio-level capital modelling. For senior executives, there is a clear implication that credit business requires a more deliberate and explicit alignment across underwriting, pricing, exposure analysis, capital parameterisation, validation, and governance.  

Only by bridging these disciplines can the entity ensure that its internal model is not just technically robust, but also strategically relevant and useful for decision making. In this session, we will:  

• Explore key methodology and parameterisation challenges that arise when modelling credit business within an internal model 
• Examine how the inherently cross-functional nature of credit risk demands closer integration between actuarial, underwriting and risk management teams than is typically seen on other lines of business 
• Discuss what organisational alignment across the credit risk value chain looks like in practice. 

This session will be in the form of an interactive panel aimed at C-suites, pricing and capital modelling actuaries at all levels, and risk managers working within entities that have significant credit exposure. 

Assumed knowledge level: 

No specialist knowledge required

Speakers:

Cherry Chan, Deloitte MCS
Naveet Teeluck, Deloitte MCS
Raymond Zhu, Deloitte MCS

Sharing professional experiences working under a South African general insurer that went through distressed times, including curatorship and liquidation. Unpacking the sequence of key events, insights, learnings and unusual experiences navigating this medium size but complex listed company with multiple insurance licences. Various topics covering a range of governance areas in the business. 

Assumed knowledge level: 

No specialist knowledge required 

Speaker:

Riyadh Mayet, Nedbank Insurance 

Multi‑billion dollar liability events are already creating catastrophic losses, from mass torts linked to chemical exposure such as PFAS, to systemic litigation linked to technology, regulation and social trends. However, unlike natural catastrophes, these risks are rarely modelled explicitly and often emerge as unexpected accumulations within portfolios. 
 
This session provides a practical introduction to casualty catastrophes, focused on helping actuaries recognise and diagnose risks within portfolios.  
 
We establish a clear definition of a casualty catastrophe and illustrate how exposures can build across business lines, products and time horizons. We then explore why progress in modelling has lagged behind natural catastrophe modelling, and identify which elements of established catastrophe frameworks should or should not be transferred to the casualty context. 
 
Finally, we will set out a structured, decision-focus approach tailored to low data but high uncertainty environments. This will help organisations identify where casualty catastrophe exposures may be building before a loss emerges, and shows how these insights can be used to track accumulation over time.  
 
This session is designed for actuaries across all functions as a standalone introduction, requiring no prior knowledge of catastrophe modelling. By the end of the session, attendees will be able to define and recognise casualty catastrophes and their drivers, identify where these risks may already exist within their portfolios, describe practical approaches to tracking and monitoring accumulation, and initiate informed discussions within their organisation about emerging casualty catastrophe risk.  

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Julien Masselot, Howden IAL
Mia Collier, Howden IAL
Jesse Nickerson, Moodys 

Most reserving teams know where the pain points are: slow data processing, manual steps, review bottlenecks, heavy reporting packs and increasing pressure to do more with less. But how do these issues compare across the market, and what are leading teams doing differently? 
 
In this session, we will share emerging findings from around 15 end-to-end reserving process reviews, focusing on what makes a process efficient, controlled and decision-useful. 
 
We will cover four areas: 

• Benchmark findings: peer ranges on cycle time, effort, model count, review layers, output burden, automation levels and data corrections 
• Good practice patterns: what stronger processes have in common, including data readiness, review design, MI quality, challenge and post-cycle learning 
• Actionable playbook: practical improvements by symptom, from quick wins to more structural changes across data, models and systems 
• AI adoption map: examples of how teams are using automation, dashboards and AI-supported tools today, where quick wins may exist, and the practical considerations around ownership, review and controls. 

The session will be practical and discussion-led rather than technical. Attendees will leave with a clearer view of how their reserving process compares to peers’, where inefficiencies typically arise, and what actions can improve quality, efficiency and governance. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Muhammad Ali, Howden IAL
Wan Heah, Howden IAL 
Joe Spicer, Howden IAL

We will articulate how a soft market impacts actuaries. Unusually, this won't focus on the technical perspective, but will incorporate the softer side of behaviours and challenges as an individual and team.  

We will draw on the (very) many years of experience we have of going through a number of soft markets in a variety of companies to help give an indication of what to expect, and possibly how best to get through the next few years. This is a matter of survival! 
 
Not for the faint-hearted. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Electric vehicle (EV) registrations in the UK accounted for 23.4% of new car sales in 2025, so understanding their insurance implications has never been more crucial. The rapid rise in Chinese EV imports is also introducing new questions around repairability, parts availability, vehicle data and future claims experience. 
 
We will provide quantitative and qualitative insights into how EVs compare with internal combustion engine (ICE) vehicles, which will be of benefit for pricing, reserving and capital actuaries. 
 
We have engaged extensively across the market, including with Thatcham Research, brokers specialising in electric HGVs, insurers and MOT data providers, and have contributed to discussions with the Department for Transport on issues relevant to the EV landscape. 
 
We will use these insights to explore quantitative and qualitative differences between EV and ICE vehicles, such as: 

• How claim frequency and severity vary 
• Differences in repair costs, parts availability, and repair times 
• The role of driver demographics in shaping observed claims experience 
• Emerging differences in claims experience between newer and older EVs 
• The impact of newer market entrants, including Chinese-manufactured EVs 

The session also touches on forward-looking considerations for insurers, including: 

• The impact of pay-per-mile electric vehicle tax 
• Potential cyber threats unique to EVs 
• The value of battery health and other EV-specific data in evolving risk assessments  

Ian Johnston, CEO of Osprey Charging, will share insights highlighting how insurers can collaborate more effectively across the wider EV ecosystem. Thatcham will also provide perspectives, drawing on their expertise in vehicle safety, repairability and emerging technologies. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Mohammad Khan, Electric Vehicles Working Party
Peter Mathieson, Electric Vehicles Working Party
Andy Goldby, Electric Vehicles Working Party

The Capital Research Group is an IFoA forum for the leading capital modelling experts in the industry.  

This panel session will explore key issues facing capital modellers across the profession. Audience engagement is encouraged.   

Issues to be discussed include: 

• AI and agentic AI in capital modelling 
• Modelling geopolitical risk and events 
• Modelling the soft market and modelling across the cycle 
• Disseminating capital knowledge in the world of remote work 
• Modelling reserve risk with a fresh eye. 

Assumed knowledge level: 

Prior knowledge required (ensuring a delegate working in that area learns something new)

Speakers:

Howard Dale, Dale Consulting
George Lewkowicz, WTW
James Toller, Beazley
Uma Divekar, Lloyd's
Jonathan Bilbul, AIG

Speaker

Richard Chalk, Bank of England

The tech sector is rapidly recognising the unique value actuaries bring as insurance subject matter experts who possess deep business acumen and the confidence to navigate massive datasets. 

This panel session explores the expanding footprint of non-life actuaries in non-traditional tech roles. Drawing on insights from previous successful Casualty Actuarial Society sessions interviewing actuaries at Google, Uber, Amazon, Expedia, DoorDash and others, this discussion pulls back the curtain on the actuarial function within big tech.

Panels will detail how actuaries optimise rideshare pricing, value gig-economy risks, manage corporate risk portfolios and proactively mitigate risks. 

We will contrast the fast-paced, iterative work culture of tech with traditional insurance environments, highlighting how actuaries pivot from regulatory compliance to product innovation.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Zora Law, Airbnb
Gary Adamson, Apollo
Elaine St. Germain, INSHUR 

The introduction of the Bank of England’s Supervisory Statement SS5/25 marks another important step in the evolution of climate risk management within the UK insurance sector. 

While climate considerations are now firmly embedded in regulatory expectations, firms continue to vary significantly in how responsibilities are interpreted and implemented, reflecting differences in size, business model and organisational structure.

This session presents findings from a cross-market survey examining how climate change responsibilities are currently managed across the industry. The survey explores governance, modelling approaches, ownership of climate risk, integration across actuarial functions, and the evolving role of actuaries under SS5/25. 

Responses will be analysed across insurer sizes and actuarial disciplines to identify areas of convergence, divergence and ongoing uncertainty.

Rather than focusing solely on regulation, the session aims to provide a practical reflection of current market practice. Key questions include:

• Which actuarial teams are leading climate-related work?
• How embedded are climate considerations in day-to-day actuarial decision-making?
• Are firms treating SS5/25 primarily as a compliance exercise or a strategic opportunity?
• What challenges arise when translating climate risk into actionable actuarial practice?

The session will also consider how actuaries can contribute to climate adaptation and resilience through product innovation and other practical interventions.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Camilla Bennett, Climate Change Working Party
Lucy Thomas, Climate Change Working Party
Heidi Llanwarne, Climate Change Working Party

The results presented cover the annual update from the IFoA Motor Working Party. 

The UK motor market continues to face significant uncertainty, with evolving claims inflation, shifts in claims frequency, pressure on affordability, and wider issues affecting pricing and reserving. This session will provide an update on current motor claim trends and prominent market issues, drawing on analysis from across the working party and insights from participating insurers.

We will cover practical implications for pricing, reserving and claims teams, offering detailed summaries of claims trends for private car insurance, including own damage, third party property damage and third-party bodily injury, from the working party’s market survey.

The session is intended to give actuaries a current and practical view of the motor market, with particular focus on the issues most relevant to assumption-setting and decision-making in the current environment.

This session will encompass and continue the work of the Third Party Working Party and the PPO Working Party as well as qualitative information on the evolving state of the motor market. 

Assumed knowledge level:

No specialist knowledge required

Speakers:

Mohammad Khan, UK Motor Working Party
Sundip Mistry, UK Motor Working Party
Saloni Baid, UK Motor Working Party

Actuaries are trained to analyse uncertainty, but so often struggle to stay clear-headed when the pressure is highest. Little attention is given to how we think, decide and perform in high-stakes environments such as reserving, capital setting and stakeholder challenge.

This session is co-presented by Clare Davis (mental fitness/performance coach) and Michael Tripp (including Fidelis, First Central, Citadel, and Baptist Insurance). Michael will bring the topic to life with practical examples from judgement heavy roles and boardroom challenges, where clarity, confidence and communication matter as much as technical skill.

We’ll explore practical, actuarial-relevant questions such as:

• Why does judgement degrade under time pressure and what does modern neuroscience say is happening?
• How do you stay responsive (not reactive) when facing scrutiny from senior stakeholders?
• How do you avoid overthinking, decision fatigue and confidence wobble in uncertain environments?
• What simple techniques can be applied immediately in a committee meeting, a contentious review, or a deadline-driven cycle?

This will be a high impact, practical session rather than a theoretical one. Attendees will leave with a clear set of tools to improve clarity, decision making and communication under pressure, plus a stronger understanding of how high-performing actuaries build mental resilience without burnout.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Michael Tripp, Fidelis Underwriting
Clare Davis, Nova Associates

Traditional reserving diagnostics help actuaries assess whether reserve estimates are reasonable, but many of these diagnostics are anchored to historical development patterns. What happens when those development patterns are moving?

Recent years have seen material shifts in claims development across a number of casualty classes, driven by claims inflation, social inflation, litigation trends, mix changes and case reserving uncertainty. In these environments, diagnostics such as IBNR-to-ultimate, IBNR-to-premium and selected loss development factors remain useful, but may not always respond quickly enough to changing development conditions.

This session will introduce residual development factors (RDFs), a practical diagnostic that provides a different lens on reserve reasonableness by comparing selected development-to-ultimate with the calendar-year development environment in which claims are actually emerging. 

The session will cover: 

• What RDFs are, how they are calculated, and the intuition behind them
• What RDFs can reveal that IBNR-based diagnostics and selected LDF checks may miss
• How RDFs can be applied through practical worked examples
• Where RDFs may be useful for casualty, inflation-sensitive, Lloyd’s and London Market portfolios
• Where the diagnostic may be less reliable.

The focus will be practical rather than highly technical. Through worked examples, attendees will learn how to calculate RDFs, interpret the results, and decide whether the diagnostic could add value alongside their existing reserving toolkit.

Assumed knowledge level:

Prior knowledge required (ensuring a delegate working in that area learns something new)

Speakers:

Muhammad Ali, Howden IAL
Joe Spicer, Howden IAL

Attendees will leave with a better understanding of Lloyd's strategy, especially in relation to capital, and will understand broadly the 2027 CPG outcomes, including softening market and geopolitical risks.

Assumed knowledge level:

No specialist knowledge required

Speaker:

Mirjam Spies, Lloyd's

Most actuarial AI talks are slideware. This one is live. 

Using only open, public pricing and claims data, we put an AI agent to work in real time on the kind of analysis a pricing actuary actually does, and narrate as it happens exactly what the model is doing and why.

The centrepiece is a live profitability and loss-ratio analysis on an open motor portfolio. We watch the agent explore the data, build a frequency–severity view, surface the segments losing money, and propose and quantify rate changes with a human actuary checking and challenging every step. 

We then set the agent several further tasks running in parallel: automated data-quality and leakage triage, a challenger cross-check model, auto-drafted model documentation, and a market scan. We then review one or two outputs alongside the main analysis if time allows.

The final minutes are an honest, behind-the-scenes look at how one pricing-software firm uses AI in its own work, from unstructured-submission ingestion and an in-IDE actuarial coding assistant, to everyday use of LLMs for models, decks, spreadsheets, research and documentation. We look at what works, what doesn't, and where humans must stay in the loop.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Harry Wilkinson, hyperexponential
Jamie Wilson, hyperexponential

This presentation takes a practical, behind the scenes look at a recent Part VII transfer, explored through the dual perspectives of the insurer delivering it and the independent expert assessing it. 

Bringing together these two viewpoints, the session will highlight not just the theory, but the reality of navigating a complex transfer from initial concept through to implementation while juggling competing priorities, evolving expectations, and the occasional moment of chaos.

The presenters will share lessons learned from both sides of the table: what it takes to get a large transaction over the line, what the independent expert is really looking for, and how both parties can work together effectively (and constructively) throughout the journey.

Topics will include:

• What a Part VII transfer involves and how the process unfolds in practice
• The role and expectations of the independent expert
• The background and drivers of the transfer
• Key considerations for the independent expert, including policyholder security and service standards
• Managing requests, queries and shifting requirements along the way
• Coordinating the transfer alongside other business change programmes
• Engaging with regulators and responding to feedback
• Addressing accounting impacts and ensuring operational readiness
• Challenges, pitfalls and complexities
• The importance of strong project management.

By combining insights from both perspectives, this session aims to provide a balanced, accessible and practical overview of Part VII transfers.

Assumed knowledge level:
No specialist knowledge required

Speakers:

Lauren Keenan, Allianz Insurance,
Simon Sheaf, Grant Thornton
Adrian Gilder, Grant Thornton

How do we ensure inclusion and fairness in the personal lines market?
What can we learn from other countries? Has GIPP and Consumer Duty helped increase inclusion and fairness in the personal lines market? How do we ensure inclusive and fair personal lines products for all? 

These are topics of substance with no single, easy answer, and give rise to questions such as:

• Can risk-based pricing align with affordability?
• What level of cross-subsidy is appropriate?
• How should high-risk customers be treated?
• How do we measure fair value?

In this session we will discuss the above questions and give viewpoints from multiple angles. 

We will also go through a proposal for Motor Re, and what this may have to look like to make the UK personal lines market inclusive and fair to all.

This will be a high-level session rather than a technical one. It should be of interest to actuaries at all levels with an interest in insuring a fairer transparent personal lines market.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Chris Price, Fairness and Inclusion Working Party
Julie Reynolds, Fairness and Inclusion Working Party
Francis Kihuiu, Fairness and Inclusion Working Party
Julianna Shing, Fairness and Inclusion Working Party

Tariffs, Iran war, what next? Nobel Peace prize, or another inflation spike? TACO theory may hold true for tariffs, but maybe not for armed conflict.  

Our presentation last year explored how the current US presidential volatility and ongoing wider political uncertainty might impact short- and medium-term inflation.

This year’s presentation will consider how surveyed views have changed in light of the sharp increase in US political volatility, along with the corresponding risks of recession.

Our Claims Inflation Working Party will provide survey participants’ feedback and views on: 

• Actual versus expected inflation at economic and class-specific level
• Inflation uncertainty, the expected next inflation spike and recession
• Social inflation and efforts to tackle it

Additionally, a key research focus this year that we plan to discuss is around capital modelling. How well does your capital model parameterisation reflect historical inflation, its volatility, and uncertainties relating to assessing it? Does your capital model sufficiently capture future inflation volatility? We will attempt to propose solutions to these quandaries.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Richard Stock, Claims Inflation Working Party
Cian Creedon, Claims Inflation Working Party
Will Sutcliffe, Claims Inflation Working Party
Karan Ahuja, Claims Inflation Working Party

The Dynamic General Insurance Stress Test (DyGIST) marks a shift in regulatory stress testing from static capital shocks to a dynamic, multi week ‘fire drill’ focused on firms’ crisis response, management actions and engagement with supervisors.

This panel will discuss what they have learnt from DyGIST, reflecting on the evolving scenarios, estimated losses under uncertainty, and how they articulated management actions that were logical, realistic and believable. 

The discussion will focus on what the exercise revealed about judgement, proportionality and coordination under pressure, rather than on technical modelling. It will also explore how these lessons are being embedded into crisis management and stress testing frameworks.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Michaela Buttigieg, KPMG
Susan Dreksler, KPMG
Cian O’Brien, Lloyd’s
Hasit Haria, HCC International Insurance Company
Lucie Stokoe, Zurich Insurance Company

Bühlmann-Straub credibility has been a cornerstone of actuarial practice for over 50 years. However, its assumptions fall short in commercial lines settings: a single signal-to-noise ratio across all accounts, equal treatment of old and recent experience, a flat complement.  

Most teams have developed their own fixes. But are those fixes coherent? Are they making things better, or just more complicated? And is there a principled way to know?

This session presents a framework that jointly estimates credibility weight, temporal decay, and complement, nesting Bühlmann-Straub as a special case. Results are validated on CAS Schedule P Commercial Auto data (96 companies), and confirmed in simulation. A companion paper will be available at the time of the conference.

Key questions explored include:

• How badly miscalibrated is standard Bühlmann-Straub in practice for commercial lines business?
• How quickly does claims experience go stale, and does the answer depend on account characteristics?
• Does fitting these components jointly produce materially better results than patching them one at a time?

The session is aimed at pricing actuaries and experience rating practitioners in commercial and specialty lines, though the framework is applicable wherever credibility methods are used. Basic familiarity with credibility theory is assumed.

Assumed knowledge level:

Prior knowledge required (ensuring a delegate working in that area learns something new)

Speaker:

Jake Morris, Allianz Commercial

Insurance pools play a critical role in providing capacity for risks that are difficult for the commercial market to absorb independently. 

Across APAC and Europe, pools have evolved to address a wide range of challenges including natural catastrophes, terrorism, nuclear risks, and emerging systemic exposures. However, the structure, governance, funding mechanisms, and effectiveness of these pools vary significantly between jurisdictions.

This session will explore the role of insurance pools across APAC and Europe and consider what lessons can be learned from different approaches. Key questions include:

• Why are insurance pools established, and what market failures are they designed to address?
• How do the objectives and operating models of different pools vary across APAC and Europe?
• What role should governments, regulators, insurers, and reinsurers play in supporting pool arrangements?
• How can pools incentivise the reduction of risk while providing affordable coverage for those in need?
• How should the success of insurance pools be measured?

Drawing on research conducted in their role as members of the Asia GI International Working Party, and their own personal experiences, the speakers will present a summary of major insurance pools in force in APAC and Europe, discuss the opportunities and challenges facing them today and examine how they are evolving to address different needs. 

The session will provide a high-level overview rather than a technical deep dive and will be relevant to actuaries, insurers, reinsurers, regulators, and risk professionals with an interest in market-wide risk-sharing mechanisms and public-private partnerships.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Charchit Agrawal, IFoA GI Asia Working Party
Darren Murch, IFoA GI Asia Working Party

Drawing on real-world experience from actuarial transformation programmes, this session provides a practical guide for leaders who want to deliver meaningful change in pricing, reserving and capital without creating unnecessary complexity. 

Understand the common reasons actuarial transformation programmes succeed or fail. Learn how to identify and mitigate transformation risks before they become delivery problems. Discover how small, productionised improvements can create momentum and accelerate long-term change. Learn practical techniques for managing stakeholders and maintaining engagement throughout large transformation programmes. 

Leave with a repeatable framework for planning and delivering successful transformation in pricing, reserving and capital.

Assumed knowledge level:

No specialist knowledge required

Speaker:

Jeremy Keating, Price Waters

AI tools and climate change considerations are increasingly shaping actuarial practice. As these areas continue to grow in importance, so too does the need to manage the risks they pose to the quality of your actuarial work. If you are keen to understand what the main risks are and how they are being managed, then this session will help to provide the answers.

This presentation will cover:

• How AI is being used in actuarial modelling and analysis
• How actuaries are currently allowing for climate change related risks in their day-to-day work
• Common challenges and examples of emerging good practice across these areas

This is an in person session delivered by actuaries from the Financial Reporting Council. It is suitable for all levels of technical knowledge and should be of interest to any actuary.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Matthew Myring-McCullagh, Financial Reporting Council
Katie Carmona, Financial Reporting Council

Thatcham have released vehicle risk ratings and will be sunsetting ABI vehicle groups by the end of 2027. This means that motor insurers across commercial, personal lines and even reinsurers need to rethink how they evaluate vehicle risk, as well as limit and guide policy limits for fleet policy holders, MGA and capacity provision. 

ABI vehicle groups has been a staple to understand motor insurance, so the change will mean one of the largest market-wide interruptions to motor pricing. During this presentation, Thatcham will provide an understanding of what vehicle risk rating is, the five pillars, and why they have moved on from ABI vehicle groups. 

The Pricing Research Group will host the session and provide actuarial context to pricing, reserving and capital actuaries who will require an understanding on how the fundamentals of motor insurance are going to change. 

Assumed knowledge level:

No specialist knowledge required

Speakers:

Charlotte Greenacre, Pricing Research Group
Zvi Ebert, Pricing Research Group
Hassan Khan, Pricing Research Group

The Risk Drivers for GI Capital Modelling Working Party will present its findings over the year. The session will set out the key historic drivers of risk for the industry, as well as practical tools for actuaries looking to transition their capital model from a copula to a drivers of risk structure. 

This will include setting out the data, parameters and methods for modelling key drivers including the underwriting cycles. It will also set out an example of moving from a copula-based model to a drivers of risk dependency structure, using this model of the underwriting cycle.

Assumed knowledge level:

Prior knowledge required (ensuring a delegate working in that area learns something new)

Speakers:

James Toller, Risk Drivers for GI Capital Modelling Working Party
Jonathan Bilbul, Risk Drivers for GI Capital Modelling Working Party

Agentic AI is the hottest of topics in technology terms, but its practical relevance for insurance is not generally well understood. This workshop will cut through the hype and explore what agentic AI actually means for actuaries: how it works, where it may help, where it may not, and how to start prototyping safely.

We will provide a quick and easy-to-understand introduction to agentic AI for professionals in the general insurance industry. We will explain the core building blocks such as tools, memory, orchestration, safeguards, and human-in-the-loop-control, connecting them to familiar insurance workflows such as pricing, reserves, portfolio monitoring, reporting, and more.

The second part of the workshop will focus on hands-on practice. We will demonstrate how to use existing tools to build a simple yet reliable agent prototype, and we will share this agent code presented as a learning experience for all. The demonstration will be both interesting and inspiring.

This workshop is aimed at actuaries, data scientists, and insurance industry leaders interested in the practical application of artificial intelligence. No deep technical knowledge is required. Participants will gain a clearer understanding of agentic AI, its opportunities and limitations in practical applications, and acquire a practical framework for determining which areas are worth experimenting with.

Assumed knowledge level:

Prior knowledge required (ensuring a delegate working in that area learns something new)

Speakers:

Alexey Mashechkin, Allianz Partners
Srinivasan Iyengar, Pacific Life Re
Tim Hill, Zenith Actuarial 

Why does the routine adoption of AI and machine learning in reserving still lag behind pricing? While interest remains high, reserving teams face persistent practical and technical roadblocks. 

In this interactive session, members of the Machine Learning in Reserving Working Party (MLRWP) will tackle these obstacles head-on, blending a progress update with future-gazing insights into tools that are reshaping the landscape.

We will share the MLRWP’s outputs over the past year, highlighting how our subgroups are supporting actuarial innovation. Discover how our literature review team is leveraging generative AI to automate the discovery and classification of actuarial research. 

Learn how the coding subgroup is successfully applying sequence neural networks (LSTMs and GRUs) to standard datasets, contrasting them directly against gradient boosting and GLM frameworks. 

Finally, we will demonstrate how our data subgroup is using simulation tools like SPLICE to overcome the industry-wide shortage of open-source data.

Moving beyond theory, we will explore the practical barriers holding back widespread corporate adoption and look ahead to emerging technologies that could finally bridge the gap between research and production.

This session will feature a presentation of our findings followed by an open Q&A forum. It is designed for actuaries at all levels, from hands-on practitioners to chief actuaries and stakeholders, who want to understand the current state of play, ask tough questions, and prepare their teams for the future of AI in reserving.

Assumed knowledge level:

No specialist knowledge required

Speaker:

Nigel Carpenter, Machine Learning in Reserving Working Party

AI presents one of the most significant opportunities the actuarial profession has faced in decades. As AI reshapes general insurance, actuaries are uniquely positioned to bring trust and accountability to its development and use. Yet as adoption accelerates, how can we best position ourselves to harness its potential while helping ensure it is implemented effectively and responsibly?

In this presentation the Professionalism, Regulation and Ethics Working Party for Actuaries Working in Data Science and AI will demonstrate how existing actuarial strengths and capabilities map directly onto key AI challenges such as explainability, bias, and regulatory compliance. They will also examine how actuaries can position themselves to make a meaningful practical contribution based on their core skillset, while continuing to develop the capabilities needed for an evolving AI landscape.

Attendees will take away practical tools and frameworks, including tools for AI model validation, approaches to assessing bias and fairness, and guidance on applying existing actuarial standards in AI contexts. The presentation will also include case studies illustrating actuarial contributions across the AI lifecycle.

Drawing on cross-professional learning, we'll set out how the actuarial profession can strengthen AI oversight and governance – positioning actuaries as key leaders in responsible, well-regulated AI.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Geraldine Finucane, Professionalism, Regulation and Ethics Working Party
Disha Dagli, Professionalism, Regulation and Ethics Working Party

This is a working party update to share the findings of the research conducted since the working party was formed, with practical use cases for the application of vines in actuarial work.

Dependencies between variables are a key determining factor of the diversification benefit in a portfolio of risks. Hence, they are a material driver of regulatory capital and should form the foundation of any portfolio level analysis. But they are notoriously difficult to describe to stakeholders, hard to measure and to implement in mathematical models.

A vine is a graphical representation of a dependence structure, on the one hand showing very clearly how different variables are connected, and on the other hand having a rich and flexible mathematical formalism, allowing for unique application to a wide range of different dependence structures.

Vines can be used to model dependencies between internal and external models, a growing area of importance given interactions between physical climate risks and related financial risks.

Any correlation matrix can be translated into a vine, and we may then choose any copula to describe the pairwise dependence between any two parts of the vine. This allows easy mixing of elliptical copulas and Archimedean copulas.

Using a vine, it is trivial to complete partially specified correlation matrices, or build arbitrarily large positive-semidefinite correlation matrices from scratch. Given a rank correlation matrix, a vine can be used to find the minimum information copula, and used to sensitivity test the stronger assumption of a Gaussian or t-copula.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Tim Harrison, Actuarial Applications of Vine Copulas Working Party
Paul Hewson, Actuarial Applications of Vine Copulas Working Party

AI tools and climate change considerations are increasingly shaping actuarial practice. As these areas continue to grow in importance, so too does the need to manage the risks they pose to the quality of your actuarial work. If you are keen to understand what the main risks are and how they are being managed, then this session will help to provide the answers.

This presentation will cover:

• How AI is being used in actuarial modelling and analysis
• How actuaries are currently allowing for climate change related risks in their day-to-day work
• Common challenges and examples of emerging good practice across these areas

This is an in person session delivered by actuaries from the Financial Reporting Council. It is suitable for all levels of technical knowledge and should be of interest to any actuary.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Matthew Myring-McCullagh, Financial Reporting Council
Katie Carmona, Financial Reporting Council

Thatcham have released vehicle risk ratings and will be sunsetting ABI vehicle groups by the end of 2027. This means that motor insurers across commercial, personal lines and even reinsurers need to rethink how they evaluate vehicle risk, as well as limit and guide policy limits for fleet policy holders, MGA and capacity provision. 

ABI vehicle groups has been a staple to understand motor insurance, so the change will mean one of the largest market-wide interruptions to motor pricing. During this presentation, Thatcham will provide an understanding of what vehicle risk rating is, the five pillars, and why they have moved on from ABI vehicle groups. 

The Pricing Research Group will host the session and provide actuarial context to pricing, reserving and capital actuaries who will require an understanding on how the fundamentals of motor insurance are going to change. 

Assumed knowledge level:

No specialist knowledge required

Speakers:

Charlotte Greenacre, Pricing Research Group
Zvi Ebert, Pricing Research Group
Hassan Khan, Pricing Research Group

The Risk Drivers for GI Capital Modelling Working Party will present its findings over the year. The session will set out the key historic drivers of risk for the industry, as well as practical tools for actuaries looking to transition their capital model from a copula to a drivers of risk structure. 

This will include setting out the data, parameters and methods for modelling key drivers including the underwriting cycles. It will also set out an example of moving from a copula-based model to a drivers of risk dependency structure, using this model of the underwriting cycle.

Assumed knowledge level:

Prior knowledge required (ensuring a delegate working in that area learns something new)

Speakers:

James Toller, Risk Drivers for GI Capital Modelling Working Party
Jonathan Bilbul, Risk Drivers for GI Capital Modelling Working Party

Agentic AI is the hottest of topics in technology terms, but its practical relevance for insurance is not generally well understood. This workshop will cut through the hype and explore what agentic AI actually means for actuaries: how it works, where it may help, where it may not, and how to start prototyping safely.

We will provide a quick and easy-to-understand introduction to agentic AI for professionals in the general insurance industry. We will explain the core building blocks such as tools, memory, orchestration, safeguards, and human-in-the-loop-control, connecting them to familiar insurance workflows such as pricing, reserves, portfolio monitoring, reporting, and more.

The second part of the workshop will focus on hands-on practice. We will demonstrate how to use existing tools to build a simple yet reliable agent prototype, and we will share this agent code presented as a learning experience for all. The demonstration will be both interesting and inspiring.

This workshop is aimed at actuaries, data scientists, and insurance industry leaders interested in the practical application of artificial intelligence. No deep technical knowledge is required. Participants will gain a clearer understanding of agentic AI, its opportunities and limitations in practical applications, and acquire a practical framework for determining which areas are worth experimenting with.

Assumed knowledge level:

Prior knowledge required (ensuring a delegate working in that area learns something new)

Speakers:

Alexey Mashechkin, Allianz Partners
Srinivasan Iyengar, Pacific Life Re
Tim Hill, Zenith Actuarial 

Why does the routine adoption of AI and machine learning in reserving still lag behind pricing? While interest remains high, reserving teams face persistent practical and technical roadblocks. 

In this interactive session, members of the Machine Learning in Reserving Working Party (MLRWP) will tackle these obstacles head-on, blending a progress update with future-gazing insights into tools that are reshaping the landscape.

We will share the MLRWP’s outputs over the past year, highlighting how our subgroups are supporting actuarial innovation. Discover how our literature review team is leveraging generative AI to automate the discovery and classification of actuarial research. 

Learn how the coding subgroup is successfully applying sequence neural networks (LSTMs and GRUs) to standard datasets, contrasting them directly against gradient boosting and GLM frameworks. 

Finally, we will demonstrate how our data subgroup is using simulation tools like SPLICE to overcome the industry-wide shortage of open-source data.

Moving beyond theory, we will explore the practical barriers holding back widespread corporate adoption and look ahead to emerging technologies that could finally bridge the gap between research and production.

This session will feature a presentation of our findings followed by an open Q&A forum. It is designed for actuaries at all levels, from hands-on practitioners to chief actuaries and stakeholders, who want to understand the current state of play, ask tough questions, and prepare their teams for the future of AI in reserving.

Assumed knowledge level:

No specialist knowledge required

Speaker:

Nigel Carpenter, Machine Learning in Reserving Working Party

AI presents one of the most significant opportunities the actuarial profession has faced in decades. As AI reshapes general insurance, actuaries are uniquely positioned to bring trust and accountability to its development and use. Yet as adoption accelerates, how can we best position ourselves to harness its potential while helping ensure it is implemented effectively and responsibly?

In this presentation the Professionalism, Regulation and Ethics Working Party for Actuaries Working in Data Science and AI will demonstrate how existing actuarial strengths and capabilities map directly onto key AI challenges such as explainability, bias, and regulatory compliance. They will also examine how actuaries can position themselves to make a meaningful practical contribution based on their core skillset, while continuing to develop the capabilities needed for an evolving AI landscape.

Attendees will take away practical tools and frameworks, including tools for AI model validation, approaches to assessing bias and fairness, and guidance on applying existing actuarial standards in AI contexts. The presentation will also include case studies illustrating actuarial contributions across the AI lifecycle.

Drawing on cross-professional learning, we'll set out how the actuarial profession can strengthen AI oversight and governance – positioning actuaries as key leaders in responsible, well-regulated AI.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Geraldine Finucane, Professionalism, Regulation and Ethics Working Party
Disha Dagli, Professionalism, Regulation and Ethics Working Party

This is a working party update to share the findings of the research conducted since the working party was formed, with practical use cases for the application of vines in actuarial work.

Dependencies between variables are a key determining factor of the diversification benefit in a portfolio of risks. Hence, they are a material driver of regulatory capital and should form the foundation of any portfolio level analysis. But they are notoriously difficult to describe to stakeholders, hard to measure and to implement in mathematical models.

A vine is a graphical representation of a dependence structure, on the one hand showing very clearly how different variables are connected, and on the other hand having a rich and flexible mathematical formalism, allowing for unique application to a wide range of different dependence structures.

Vines can be used to model dependencies between internal and external models, a growing area of importance given interactions between physical climate risks and related financial risks.

Any correlation matrix can be translated into a vine, and we may then choose any copula to describe the pairwise dependence between any two parts of the vine. This allows easy mixing of elliptical copulas and Archimedean copulas.

Using a vine, it is trivial to complete partially specified correlation matrices, or build arbitrarily large positive-semidefinite correlation matrices from scratch. Given a rank correlation matrix, a vine can be used to find the minimum information copula, and used to sensitivity test the stronger assumption of a Gaussian or t-copula.

Assumed knowledge level:

No specialist knowledge required

Speakers:

Tim Harrison, Actuarial Applications of Vine Copulas Working Party
Paul Hewson, Actuarial Applications of Vine Copulas Working Party

The insurance market cycle is one of the few things everyone agrees exists and nobody seems able to stop. 
 
As competition heats up and pricing pressure builds, pricing actuaries find themselves at the centre of critical decisions. But do they truly influence outcomes, or are market forces simply too powerful? 
 
Join us for a fast-paced competitive debate featuring sharp arguments, audience polling and plenty of challenger questions as we tackle controversial topics including: 

• Are we better prepared for today's soft market than the last one? 
• Do pricing actuaries actually make a difference? 
• Will AI and new technology reduce or amplify market cycles? 
• Has more data improved decision making, or just increased confidence? 

Come ready to take a side. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Alice Boreman, Managing the Cycle Working Party
James Anderson, Managing the Cycle Working Party

As we all know, IFoA members are bound by the Actuaries’ Code, and we’ve all heard the (occasional) tale of an actuary who has fallen foul of it. In recent years, the Code has been brought back into the spotlight, as efforts to update it were met with confusion and (in some instances) opposition.  

This begs the question – is it possible that being bound by such a code of conduct is causing actuaries more constraint than it’s worth? 
 
In this talk, Cian will explore and argue for the pragmatic merits of the Actuarial Code, leaving recent discourse to one side and drawing on real, pivotal historical events and some of the great (and more digestible) philosophers of ages past to make this viewpoint evident. 
 
This talk will draw upon Cian’s recent experience in delivering professionalism sessions to the London Market Actuaries’ Group, of which he is longstanding chair, and the wider London market. His aim is to discuss topics of professionalism in a witty and entertaining style, coupling sound argument with enthusiastic engagement.  
 
In addition to being a seasoned actuary with nigh on two decades of experience, Cian has a passion for history and will bring both actuarial acumen and his love of classics to bear in a talk that promises to inform and delight. 

Assumed knowledge level: 

No specialist knowledge required

Speakers:

Cian Creedon, AON

As machine learning (ML) reshapes insurance, the actuarial profession must lead or risk being left behind.  
 
ML is a natural extension of actuarial skillsets, yet many actuaries struggle to invest the time needed to build expertise, creating a growing gap between industry expectations and capability. 
 
We bring together representatives from the IFoA AI & Emerging Technologies Practice Board and the Machine Learning in Reserving and Actuarial Data Science Working Parties to examine how this gap can be addressed in practice. The panel combines leadership across the profession with hands-on experience of developing, applying and governing ML techniques in insurance. 
 
We will explore the 'actuarial triangle' of responsibility between members, employers and the profession, and ask: 'Who needs to do what to accelerate progress?' Together we will explore practical ways to overcome current barriers, including: 

• How actuaries can realistically build ML skills alongside day jobs 
• How employers can provide structured training time and support 
• The role of working parties as a form of applied, collaborative training 
• How the IFoA and universities can better support early-career talent and engagement. 

This is an interactive session, designed to invite audience perspectives and challenge assumptions through live discussion around what tangible changes are needed. Attendees will leave with practical ideas for how actuaries, employers and the profession can accelerate ML capability. 
 
The panel brings a combination of perspectives across consulting, industry and research, with strengths spanning reserving, predictive modelling, AI governance and interpretability, including GI leadership, MLR chair, ADS chair, and Brian Hey Prize-winning ML research. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Michael Ramyar, Karol Gawlowski and Alexey Mashechkin, representatives from the IFoA AI and Emerging Technologies Practice Board, the Machine Learning in Reserving Working Party and the Actuarial Data Science Working Party

Frequently, internal model correlation structures are too rigid, adopting a one-size-fits-all approach to capturing dependencies across classes with very different loss drivers.  

Whilst aggregate volatility may pass validation tests, one may question whether the chosen mechanisms truly reflect the underlying common causality of the frequency and severity of large losses.  

Given the competitive reinsurance landscape, do our dependency modelling choices therefore impair the assessment of value derived from multi-class excess of loss reinsurance? 
 
This case study-led session will investigate how this derived value changes under different correlation frameworks and whether the mechanisms themselves mirror reality. Focus will include: 

• Working layer versus capital protection covers 
• Underlying loss drivers for different classes of business 
• Copula versus loss driver correlation approaches 
• Frequency versus severity correlation 
• Impact of large loss attachment 
• Ordering of correlation tiering

This will be a mathematically light session, of potential interest to any actuary working in fields of capital modelling, risk or reinsurance. Some knowledge of stochastic modelling will be assumed. 

Assumed knowledge level: 

Prior knowledge required (ensuring a delegate working in that area learns something new) 

Speaker:

James Francis, Gallagher Re

Following a major pricing transformation, many insurers struggle to fully realise the value promised during implementation. Once the new models and infrastructure are live, attention often shifts quickly to business-as-usual delivery, leaving significant opportunities underutilised.

This session explores how insurers can maximise value immediately post implementation, drawing on perspectives from a pricing software provider, a consultancy specialising in pricing transformation, and industry practitioners who have recently implemented new pricing systems and infrastructure. 
 
The discussion will focus on practical lessons learned from the post-implementation phase, including:

• Prioritising the activities that unlock the greatest business value from new pricing infrastructure 
• Embedding effective governance, monitoring, and model ownership approaches 
• Identifying commonly overlooked capabilities and technical debt left behind during implementation 
• Avoiding common post-implementation pitfalls that reduce long-term return on investment 
• Demonstrating measurable commercial and operational benefits to stakeholders 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Jack Miller, Apollo Underwriting
Tokunbo Ahmed, Cincinnati Global Underwriting
James Nicholl, APR LLP
Rob Williams, hyperexponential
Morgan Smith-Woodhams, Cincinnati Global Underwriting 

PRA and Lloyds have consistently warned about the optimism embedded in business plans. This risk will heighten as the insurance markets soften – strategic business plans do frequently succumb to institutional optimism.

In an effort to maintain top-line premium volume, organisations often rely on unrealistic rate assumptions, overestimating underwriting discipline while underestimating the compounding effect of sticky claims inflation. This session, presented by the TORP Working Party, establishes best practices for constructing resilient business plans during market downturns.

We compare the structured, benchmark-driven Lloyd’s capital plan approval gates with the fast-paced, high-frequency portfolio adjustments required in retail personal lines, highlighting how both systems must solve the same core structural forecasting challenges.

We explicitly detail what goes wrong when planning loops fail: uncoordinated data handovers, lagging indicators, conflicting interests between the teams and a dangerous disconnect where pricing erosion is masked by historical reserving buffers. We will provide a blueprint to optimise cross-functional interactions for planning. 

The presentation focuses on calibrating critical, high-sensitivity assumptions and how to get them right by using forward-looking data, expert judgment and feedback loops. 

How can today’s cutting-edge data technologies and AI help in facilitating these processes?

Building on this, the presentation covers a case study of how the reserving actuary can evolve into a strategic lynchpin for planning and an adaptive leader driving change. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Arun Vijay, TORP Working Party
Param Dharamshi, TORP Working Party

This session will provide an update on the Prudential Regulation Authority’s (PRA) view of the current general insurance risk landscape, including its plans for 2027. The session aims to provide a clear, practical overview of current and upcoming regulatory developments and insights from recent thematic work. It will equip attendees with the context needed to plan effectively and engage proactively with the PRA’s evolving agenda. 

Assumed knowledge level: 

No specialist knowledge required 

Speaker:

Daniel Curtis, Bank of England

The PRA didn't ask insurers to plan for failure. They asked them to plan for an orderly exit, and in doing so, exposed something far more uncomfortable: how few firms truly understood themselves. 
 
Solvent Exit Analysis (SEA) forced insurers to answer questions they'd never had to face. Who actually owns the relationship with your third-party administrator when you're winding down? If your intercompany reinsurance unravels, in what order do the dominoes fall? How much capital do you really need to keep the lights on while handing back policies? 
 
Across firms ranging from lean monolines to sprawling composites, we saw wildly different answers and approaches. Some spent £50,000, others £500,000. Some were done in eight weeks, others were still going at six months. Some drew heavily on their ORSA, hoping for the best, others dragged documentation from different business functions, reconciling it for the first time. 
 
What made the difference wasn't size or sophistication. It was a firm’s ability to construct a coherent narrative. One that held up under scrutiny, threaded together dependencies, and gave the board genuine confidence. 
 
This session draws on extensive experience supporting a vast and diverse range of firms through SEA development. We'll share what separated the plans that genuinely worked from those that merely complied, and what those lessons mean for embedding strength and adaptability into its business-as-usual thinking. This session will be insightful for any actuary interest in seeing how resilient firms weather the storm. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Oscar Marshak, Grant Thornton UK
Frederick Akinwunmi, Grant Thornton UK

The actuarial control cycle (ACC) still works – just not everywhere. And in the places it doesn't, the risk isn't that it fails visibly. It's that it fails silently. 
 
No need to attend this session if you strongly believe: 

• Most actuarial problems are fundamentally stable, ordered, and optimisable – and will stay that way 
• When your forecasts are wrong, the right answer is always “go deeper” (more data, more sophistication, more refinement) 
• AI/ML will behave like normal code – transparent, explainable, and consistent through time – validate once and monitor simply 
• Uncertainty is mainly a deficit of technique, not a property of the system 

If any of those feel shaky, this session is for you.  

We’ll propose the living control cycle (LCC) as an extension (or successor) to the ACC, aimed at work that sits in a more entangled, dynamic space. We will set out the case for change, the core ideas, and a set of practical starting moves actuaries can try immediately in reserving, pricing, and portfolio management. We will illustrate these with examples from the Better Sensing and Responding to Change Working Party’s research on unstable environments and surprise detection, rather than theory alone. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Christopher Smerald, Better Sensing and Responding to Change Working Party
Ian Thomas, Better Sensing and Responding to Change Working Party
Gregor Fabjan, Better Sensing and Responding to Change Working Party

AI and machine learning are now firmly embedded in the general insurance pricing and underwriting agenda, and the focus is rapidly moving on from discussion and experimentation to BAU implementation and the need to demonstrate real value.   

But adoption remains uneven, management expectations can stray into the realms of fantasy, and the path to successful implementation is still not fully charted. 
 
In this session, the IFoA AI/ML for GI Pricing Working Party will cut through the hype to provide an evidence-based view of how AI is being used in practice across the industry – from personal lines to London market – highlighting where it is working well, and where it is not.

Drawing on a 2026 cross-market survey of actuaries and pricing practitioners, alongside insights from the working party’s research streams, we examine: 

• The current state of play: where AI is delivering genuine value today and where expectations continue to outpace reality 
• The most significant barriers to adoption 
• How priorities and maturity differ across market segments – particularly between predictive modelling for personal lines, and the use of generative AI to support underwriting and decision-making in the London market – are shaping divergent adoption paths across the industry 
• The implications for actuarial teams, including skills, operating models, and areas for investment over the next phase of adoption. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Ed Plowman, IFoA AI/ML for GI Pricing Working Party
Zvi Ebert, IFoA AI/ML for GI Pricing Working Party
Dylan Liew, IFoA AI/ML for GI Pricing Working Party

Flood experience in the UK has been highly variable and high profile. Flood Re was set up in 2016 to subsidise flood risks in the UK and make flood insurance available and affordable. Flood experience in the UK has been less adverse than feared immediately after the 2007 flood losses, with investment in flood defences being instrumental in mitigating losses. However, considerable uncertainty remains as to the impact of climate change. 
 
This session will: 

• Cover historical flood losses in the UK 
• Discuss the recent advances in flood modelling, both from vendor models or academics 
• Share some insights into Flood Re's own experience of flood losses 
• Socialise best practice in flood modelling, drawing on the specialist and wide experience of practitioners in industry.   

This will be a panel discussion with specialists, and open to audience dialogue for maximum benefit to attendees. 

Assumed knowledge level: 

Prior knowledge required (ensuring a delegate working in that area learns something new) 

Speakers:

David McNamara, Flood Re
Krish Kistnassamy, Howden Insurance Consulting

General insurance actuaries are facing an expanding set of complex, interdisciplinary challenges, from inflation and affordability to ESG, climate‑ and nature‑related risks, AI and data science, and evolving regulation and professionalism expectations. Yet many of these issues are still discussed in specialist silos, and traditional conference formats can make it hard for actuaries to explore how other people’s hot topics affect their own work. 

This highly interactive session uses a bespoke expansion of “Cards Against the Actuary”, a game created and successfully piloted in an IFoA regional community, to surface and discuss these themes in a playful but purposeful way. The original game generated thousands of social media engagements and significantly increased participation in online dialogue among actuaries, demonstrating its ability to lower barriers and stimulate conversation. 

The GIRO edition will feature new cards based on current GIRO priorities (for example climate modelling, AI in reserving, solvency exit, geopolitical risk) and related thought‑leadership such as the future of insurance and planetary solvency work on climate‑nature‑society linkages. Through structured gameplay and facilitated debriefs, participants will explore serious ideas, hear diverse perspectives from peers, and leave with a richer understanding of challenges both within and beyond their own GI niche. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

Mohamed Elsheemy, Crystallise LTD
Victoria Moffett, NFU Mutual

Most firms operate some form of fast close¹ reserving process, allowing quick monthly/quarterly reporting of updated reserve results in a tight timeframe.  

A key limitation of the process is that the reserving approach is usually fixed for each reserving class, with limited time available to validate² and update it during the process.   

We present a data- and diagnostic-driven framework that reserving teams can set up and apply quickly on roll-forward. This framework incorporates both traditional metrics and machine-learning based diagnostics to help actuaries make smarter roll-forward decisions in less time. This allows for targeted, light touch adjustments to the reserving process in a logical, structured and potentially automated manner.  

Our framework incorporates:  

• A new machine learning-based reserve deterioration score, which we present in detail 
• Traditional diagnostics such as AvsE, analysis of change, graphical metrics such as incurred and paid as percentage of ultimate claims – we summarise those which efficiently provide the most complete picture 
• A mechanical evaluation and prioritisation process based on the above, which can be fine-tuned to the governance processes and reserving culture of specific companies. 

We set out a worked example of how to apply this framework reserving data and use it to guide decision making, allowing actuarial teams to make targeted adjustments to reserves in a time-efficient manner.  

• ¹IFoA ROC Working Party: The fast close process
• ²IFoA TORP Working Party: Your reserves may be best estimate, but are they valid

Speakers:

Ed Harrison, LCP
Charlie Stone, LCP

The Managing the Cycle Working Party have developed a toolkit for pricing actuaries working in the Lloyd's and London markets. As the market continues to show signs of softening, we will set out key focus areas for pricing to actively contribute to navigating the market conditions. 

Assumed knowledge level: 

No specialist knowledge required 

Speakers:

James Anderson, The Managing the Cycle Working Party
Alice Boreman, The Managing the Cycle Working Party
Subaigh Prem, The Managing the Cycle Working Party