This piece is by Daren Rudd, Head of Business, Technology and Innovation Consulting, Insurance UK, CGI.
It starts with value
In the world of insurance, transformation is often driven by the search for cost reduction. Whether through automation, digitisation, or restructuring, organisations have historically pursued operational efficiency as the primary goal. However, true transformation needs be about more than just cost-cutting, it must create real long-term value for the business, its people and its customers. This critical point is often lost in the excitement of new technology, and this is especially true with the current buzz around generative AI (GenAI).
Value in transformation needs to go beyond short-term financial gains. It requires a strategic approach that enhances customer experience, empowers employees, and future-proofs business models. To strike this balance, insurers need to critically evaluate how they integrate technology while preserving the human elements that have long defined the insurance industry.
AI: The elephant in the room
AI is dominating boardroom conversations, with insurers heavily researching its potential. GenAI and large language models (LLMs) in particular have captured industry attention. However, while AI, in its various forms, offers opportunities to improve business operations, it is not a silver bullet that will solve every challenge by itself.
We can draw parallels with the industry’s adoption of cloud computing. Many insurers transitioned to cloud-first models without adequately re-engineering their applications for cloud, leading to missed expectations on cost and scalability benefits. Similarly, layering AI onto legacy processes, data, and systems will not automatically deliver long-term gains.
Instead, insurers need to be pragmatic in finding suitable AI point solutions, driven by solid business cases, rather than assuming it will be a universal solution for all problems. It is also important not to get swept-up by the overall excitement over GenAI and LLMs. They are just one subset of AI capabilities, so it is important to assess the problem to be solved and then look at the right tool, which may not even be AI.
Change teams will need to manage the business tendency to want a ‘tech silver bullet’ which often overshadows the hard work of systemic change, and what this will mean in practice for the people impacted across the organisation.
The real work of transformation
AI should not distract from fundamental organisational and legacy transformation efforts. Our Voice of the Client (VOC) analysis showed that 40% of digital investments for our insurance clients’ were focused on AI projects, while only 4% was being directed towards legacy modernisation.
However, nearly 80% of our clients still cite legacy systems as a key constraint to delivering digital transformation. This is a worrying disconnect and suggests that AI is being seen as a one-stop solution without considering the other fundamental, and frankly difficult, change that will be required.
If insurers truly want to leverage AI effectively, they must address underlying legacy infrastructure and data challenges at the same time. Without modern, adaptable systems, AI implementation risks being another costly overlay fundamentally handcuffed from being able to deliver the expected business transformation.
Human in the loop: rethinking roles
While there are opportunities from the new technology, it is important to remember that it is still evolving at pace. Questions remain as to its effectiveness and reliability, especially when we consider the regulated nature of the insurance market. The biggest uncertainty however is the impact on the industry’s most valuable asset, its people.
AI’s integration into insurance workflows must be carefully managed to avoid unintended consequences. One major risk is reducing human roles to ‘error capture’ functions. If employees are merely validating AI-generated outputs, especially if there are still questions on the reliability of the response, their work may become unfulfilling, leading to disengagement and loss of expertise.
It is vital that an AI strategy should consider not just how it can be used to replace work already done, but instead how it can be used to augment human capabilities and change what is possible. An obvious starting point is where AI can enhance decision-making rather than simply automating a process.
Rather than following the previous pattern of earlier automation technologies which were focused on digitising an existing process, we should ask how can we instead redesign roles so that they leverage AI without diminishing professional expertise?
Additionally, insurers must be mindful of long-term workforce development. If decisions are increasingly deferred to AI models, how will junior underwriters, for example, gain the experience needed to become future senior underwriters? The industry must ensure that today’s professionals still have meaningful learning opportunities to develop into tomorrow’s leaders.
The changing cost dynamics of AI
The rapid pace of AI innovation introduces another layer of complexity. We have seen in recent weeks the existing cost paradigms for AI training have already shifted significantly, with the release of new models such as DeepSeek reshaping expected model training costs.
It is also important for decision makers to consider that the current cost models for LLMs are not sustainable in the long run. This means today’s ROI calculations may not hold in the future. Any strategy must consider the likelihood of cost corrections in AI consumption costs and how that will impact long-term business cases.
The likely reality is that initial efficiency gains may be offset by increasing consumption, maintenance and adaptation costs, even with the introduction of new more efficient models and chip infrastructure. ROI models should be looking at several scenarios for how those costs may change over time.
Balancing efficiency with long-term workforce development
Ultimately, insurers must balance the drive for initial cost reduction and efficiency improvements with broader strategic imperatives. Efficiency is important, but not at the expense of skills development and building meaningful career pathways.
There is already a resource challenge within the industry. Industry analysis showed that the number one constraint for achieving business priorities was talent shortage, sitting above even budget constraints.
The correct application of AI could support mitigating that impact, but it also risks damaging the attractiveness of insurance by reducing meaningful roles in early to mid-careers. This must also be balanced with managing organisational resilience with over-reliance on AI-driven automation risking eroding internal expertise and outsourcing key competencies to third parties.
AI presents incredible opportunities, but requires thoughtful implementation aligned with long-term value creation for both business and people. By focusing on both people and technology, insurers can ensure their transformation efforts deliver not just efficiency, but sustainable, human-first progress.
Be the first to comment