The future of data and AI in delegated authority

Eames Partnership hosted an exclusive delegated authority breakfast roundtable on Wednesday 20th November 2024. I was joined by heads of delegated authority, heads of operations, chief underwriting officers, consultants and advisory board members to discuss the likely impact of emerging technologies on delegated authority, with a particular focus on generative AI. 

The latest trends in delegated authority 

Eames interviewed nearly 100 people from the top 20 Lloyd’s syndicates and wider general insurance market in 2024, supplementing their insights with a series of client surveys to understand what’s happening in the market right now.  

  • Delegated is growing at a far greater pace than open book – The majority of syndicates have more than 50% of their business distributed on a delegated basis. Delegated currently represents over 40% of the volumes at Lloyd’s and over 10% of the UK’s general insurance market. 

  • The potential of delegated authority – Historically, DA has been seen as the poor cousin of open book, but it can provide an opportunity to innovate, test new products and move to portfolio underwriting, allowing one underwriter to control more books of diverse risks. 

  • Bordereaux the number one issue – Data challenges coming a close second. 

Current challenges in delegated authority 

Eames’ online polls identified data challenges (selected by 58% of respondents) as by far the biggest challenge in delegated authority right now, with talent availability the next most prominent challenge at a distant 19%. 

While 72% of managing agents and carriers are interested in AI for process automation, many are unclear on what to use generative AI for, or even what it is. Data availability (55%) is the most significant barrier to incorporating generative AI in delegated authority, followed by lack of internal expertise (27%), as experts’ knowledge is not properly shared and often leaves the business when they do. 

Ethical challenges and lack of regulation (both rated as highly challenging by 52% of respondents) are the biggest challenges in implementing AI. 

Discussion 1: The power of emerging technologies 

Data is massive for organisations right now, particularly when it comes to maximising AI, but the general insurance industry is not getting the most out of its data. Why is this? 

There’s no such thing as ‘perfect’ data 

With data silos and varying data standards, organisations could look at AI automation even when the data isn’t in perfect order. Perhaps ‘good enough’ data is a sufficient base for underwriters to work from, while the organisation accepts a degree of risk when devising their strategy. 

Inconsistency 

There are no common industry data standards in delegated authority. Common data taxonomies and a shared data entry system would ensure consistency of formatting and understanding between all carriers, cover holders and MGAs. This would make carriers more confident about the quality of the data and reduce the time and resource currently spent on processing, storing and maintaining data that would be better spent on data analysis. 

Working out the business case 

While some organisations are struggling to understand the potential uses and returns of AI, others are benefiting from small scale AI usage. For example, inadequate or siloed data is less of a problem when looking at speciality commercial insurance, where large volumes of data are not necessary for accurate risk assessment. Some providers are working with amalgams of frontier AI models to deliver specific projects in customer service representation. 

C-suite understanding 

Boards often don’t have a clear understanding of data, digital technologies or even delegated authority itself. This means the C-suite don’t know what data is available and often ask heads of delegated authority to take on too much, without the resources, talent or support they need. 

How do we mitigate these? 

Collaboration 

Underwriters, operations and data scientists working effectively together to solve the same problems could be more important to success than new technologies.  

Asking the right questions 

Before coming up with solutions, organisations need to take a step back and decide what problems they’re looking to solve. Only then can they understand what kind of data they need and what they should aim to get out of it. If AI is the answer, what is the question? 

Discussion 2: People, talent and skills development  

There’s a lack of cohesion in organisational structure within delegated businesses, with 77% of DA teams reporting into operations (as firms look to increase operational efficiencies and reduce the risk of data leakage) and 23% reporting into underwriting. 

This isn’t helped by a lack of guidance or sharing of best practice on structuring and managing delegated authority business. With varying degrees of success, DA businesses often outsource and offshore their administrative functions before considering whether AI or other technologies could help. 

Overcoming the data challenges facing DA  

Exacerbated by delegated authority and operations being mostly separate and reporting into different board members, the three main organisational and operational challenges facing delegated authority executives are lack of connectivity, lack of communication and lack of collaboration. How can DA use data more successfully? 

  • Data scientists – An organisation requires skilled and trained data scientists to manage its data successfully. However, these professionals also need to be properly incorporated into the business to understand the idiosyncrasies of bordereaux and delegated authority. 

  • Collaboration – The underwriting, operations and data teams need to work together to understand what problems DA needs to solve. At the same time, the nature of DA business is evolving and data is becoming more important, so heads of delegated authority need to have strong data awareness. 

  • A change in attitude – DA could use its specialty and quirky nature to attract new talent. The challenge here may be overcoming C-suite apathy and making a strong business case for talent, culture and alternative organisational structures. 

  • Using the latest technology – Working with the latest tech innovations, such as automation, will interest the next generation of talent and give them the opportunity to learn, as well as fuelling a culture that encourages innovation. 

  • A focus on data – Cultivating a dynamic data-driven environment will strengthen the organisation’s appetite to solve the data challenges of DA. 

  • Repositioning – Highlighting that operations can be exciting to work in could make DA more attractive to potential candidates who would otherwise find underwriting more appealing. 

Delegated authority – no longer the poor cousin? 

Delegated authority faces a number of people, talent and skills challenges when it comes to emerging technologies – but these are by no means insurmountable. With greater organisational focus on DA business generally, and conscious effort from DA itself to shout about the things that mark it out from the traditionally more appealing open book, delegated authority has the opportunity to step out of the shadows – and data and AI can be a key part of that. 

These insights draw upon a significant body of research comprising client interviews and survey data, gathered at the end of 2024. If you’d like to discuss these in more detail, please contact me at ben.johnson@eamespartnership.com and I’d be happy to run through our findings. You can download a full report on the discussion ‘What’s the future of data and AI in the delegated authority space?’ below. 

Download Report

Fresh Content

Explore our latest market insights, career advice, and more.

The future of data and AI in delegated authority
Accessibility and Languages