Practical toolkit for AI adoption in the Lloyd’s Market

AI is now moving beyond experimentation into everyday use, with many organisations now adopting a structured deployment model, with appropriate governance, to bring AI into their products, services and solutions.

It’s a significantly evolved landscape from two years ago, when we first partnered with Lloyd’s Market Association (LMA) to deliver an in-depth survey into AI in Risk Management. We’ve revisited this survey to capture today’s outlook on the accelerating adoption of AI in the Lloyd’s market. While the results show a dramatic shift in AI usage across different insurance functions, they also show a much greater awareness of the risks and opportunities that lie ahead. 

You can view the survey and the accompanying toolkit here:  

• Access the toolkit
• Read the survey

The AI adoption toolkit has been designed as a guide for LMA members who are developing or enriching their existing AI governance frameworks to ensure the responsible, transparent and effective use of AI systems across their operations. It has been developed using insights from our recent survey on “AI Risk Management”, bringing together industry insights from Chief Risk Officers (CROs) and Chief Operating Officers (COOs) amongst the respondents. We collated insights from 39 participants, representing over 60% of the market’s stamp capacity. Deeper perspectives were provided by 11 in-depth interviews. 

In the resulting report, we introduce five principles that LMA members can consider when designing their own AI policy or framework: 

• Governance and accountability 
• Risk tiering 
• Data protection, security and intellectual property 
• Training and awareness 
• Pragmatic adoption 

The report also provides practical templates and summarises existing guidance from other organisations such as the ABI, UK government, EIOPA and NIST. 

AI risk management – a Barnett Waddingham perspective 

In analysing the outputs of the survey, our experts have been encouraged by how much progress we have seen in this space, while recognising that widespread adoption remains some years away. It is clear very few firms can, or do, claim to be at the forefront of the AI revolution, with most firms still in the early stages of thinking and experimenting. However, a large proportion of survey respondents have an AI framework in place and are actively managing the risks around AI systems and tools. This provides a good foundation for future AI scaling within businesses. 

We have also observed that companies with a US presence tend to be more advanced in terms of their adoption relative to other players in the market. In our view, there are two key reasons for this: 

• These groups have more financial capital to invest in AI systems, and deeper pockets play a key part when it comes to investing in new technology. Having these financial resources enables these players to try, fail and learn quickly until they find an appropriate solution to their requirements. It also makes sense from an economies of scale perspective, as solutions can be deployed quickly and consistently across the US and beyond. 
• These groups generally operate in more than one developed territory. As these territories, such as the UK and EU, have more established AI regulations, they provide a clearer blueprint for AI rollout across wider organisations. 

Our research shows an additional factor determining the speed of AI system roll-outs is personal drive from the leaders. A few respondents shared that their relatively quick adoption was due to senior leaders within the business being keen to be ahead of the curve in embracing these new tools. 

Despite the optimism visible in the report, it is not without its concerns. Participants questioned whether they were investing enough to truly benefit from AI, and reflected on whether there were tangible benefits to being an early adopter in an area that is quickly evolving. We urge firms to consider both opportunity and adaptability when devising an AI strategy, which must be in line with their business needs. 

AI technologies are evolving very quickly and it’s clear there is no single right or wrong approach to an AI framework. What’s important, however, is that companies continue to balance the risk and opportunity, and devise a practical strategy and robust risk management for AI adoption. 

To discuss the findings of this survey, the toolkit, or to gain any further insights into AI Risk Management, please reach out to us or your contacts at the LMA.