AI from Potential to Performance
Jonathan Sharp, CEO, Britannic - 9th December 2025
AI is transforming the insurance industry by improving ‘Know Your Customer’ (KYC) with behavioural analytics and enhancing the underwriting and claims process, reducing operating costs and meeting increasing customer expectations. However, insurers should proceed with caution when implementing AI solutions because if it isn’t operationalised correctly mitigating all the risks it comes with and isn’t aligned with your business and people, then it is doomed to fail.
Step Up
The first step is to determine who will own the project internally. An AI project needs an owner responsible for overseeing and managing it daily to monitor performance and delivery and to identify areas for improvement. Especially in contact centres, AI projects are not finished once deployed; they are live entities that should develop and improve naturally over time. Without someone taking responsibility, the technology might remain unused and fail to deliver the expected results.
Joining up the Dots
The key to a successful technology project is connecting the technology to the business. An owner will help bridge the gap ensuring it is operationalised correctly and is adding value. Owners are responsible for the strategic planning, training and championing it, and ensuring it stays on track. Checking in constantly to see if it is achieving the goals and aligned with the workflows, customer journey and outcomes.
Laying Strong Foundations
Next, you will need to identify your pain points, the challenges you face in the customer journey, and the improvements you wish to make. It is advisable to ask your agents and customers about these issues and how they can be enhanced, as they use it daily.
From here, you can then set measurable objectives that you want to achieve. Gartner reported that only 48% of digital initiatives meet or exceed business outcomes, demonstrating the importance of preparation and ownership. Additionally, it highlights the staggering number of AI projects that are falling into the dark abyss and failing. Implementing AI should be a considered decision, not just because every other insurer is doing it and you are suffering from FOMO (fear of missing out).
A Trusted Consultant
It is crucial to work with an experienced, trusted technology consultant who can guide you through every step of the process, rather than just implementing the technology and leaving the rest to you. They can also help identify pain points in your customer journey by conducting an AI audit on your customer data and contact centre, enabling you to detect problems, some of which you may not even be aware of.
People Power
It is vital to train your employees on the AI solution; without their involvement, the project cannot succeed. AI relies on people, so it’s important to involve them from the outset, explaining why it’s needed, what the issues are, and how they believe their processes could be improved. Provide training on prompts, enhance their skills with the AI solution, and identify potential improvements.
Reassure them that it isn’t being implemented to replace their jobs but to enhance their roles, making it more interesting and fulfilling. Enabling them to delegate routine daily administrative tasks to the bots and concentrate on more complex, high-value claims that require human skills such as critical thinking, reasoning, empathy, and understanding.
Employees should also be trained to work with AI to extract value from it. Set a culture of learning and innovation that encourages staff to suggest ways to improve workflows, decision-making, and service delivery. Additionally, it is essential to enable them to report issues that aren’t working. We are all learning together in the world of AI.
Data Skills
The UK insurance market has one of the most mature workforces, with a wealth of knowledge and expertise that must be passed on to younger generations. AI is revolutionising underwriting, claims, and the customer experience, transforming how policyholders are served. However, there is a shortage of data scientists and engineers capable of extracting and interpreting data into actionable insights for the business.
Stats reveal that insurers are less confident when it comes to predictive analytics with only 13% using them highlighting the need for education, increased awareness and skills (Insurance Times AI Claims Survey (RDT).
AI analytics delves into the data, helping you understand your customers' buying behaviour better by tracking quotes that are converted and those that are not, and identifying the differences. For companies that utilise data analytics, they can offer more compelling prices and propositions, personalise offers that stand out from competitors, and capitalise on these opportunities.
Moving Closer to Customers
AI is enhancing Know Your Customer (KYC) processes in the insurance market by making them faster, more accurate, and cost-effective through automation and advanced analytics. It helps improve customer experience with quick verification and onboarding, as well as throughout their policy lifecycle. Additionally, it provides more precise risk assessment and scoring, as AI algorithms can analyse vast amounts of data from everywhere, enabling insurers to pass high-value cases to human agents.
AI solutions offer richer customer insights and analytics, enabling insurers to deeply examine their customers’ data and behaviours. This helps them understand their customers better and personalise products, pricing, and offers accordingly. It facilitates a shift from a traditionally reactive approach to a proactive one, fostering trust, loyalty, and customer retention in a competitive industry.
By personalising communications and services, insurers can differentiate themselves in the market and prioritise cases more efficiently, making customers feel more valued and cared for. While chatbots and conversational AI can provide a 24/7 service, improving efficiencies and allowing human agents to handle higher-value or more complex cases.
Clean Up Your Act
In a heavily regulated industry handling vast amounts of customer data, it is crucial that your data is AI ready. A survey revealed that 72% of UK underwriters considered fragmented or unstructured data as the main obstacle to AI transformation.
Insurers’ data must be accurate, current, and validated to ensure customer fairness and meet the FCA’s outcomes for Consumer Duty. AI will only succeed if it relies on clean and reliable data, which also helps reduce the risks of errors, misquoting, and providing incorrect information. If the data and KYC are not robust, you will not only breach regulations but also harm your reputation and lose customers.
The Stakes are High
It is essential for humans to supervise any AI solution by implementing guardrails, guidelines, and governance. This ensures data security and helps you to meet all necessary compliance and regulations.
AI is and will continue to transform the insurance industry with underwriting, claims, and distribution, but it also involves risks such as model risk, where algorithms can make mistakes, leading to inaccurate pricing or decision-making when faced with new data. This can result in conduct risk, affecting customers because it has misquoted, mis-scored, or mis-triaged, breaching the FCA’s regulations. Therefore, it is essential that companies ensure human oversight to monitor and intervene when necessary.
Guardrails and governance with AI are also necessary to manage operational risks such as poor data quality, cybercrime breaches, and outages. If not addressed properly, these issues can be detrimental to your business, undermining trust.
Bias and discrimination risks can also be evident in AI-driven underwriting and claims, where AI might make decisions that breach equality law and consumer protection regulations.
When implementing AI, companies need to ensure that it is planned strategically, considering the different risks that may arise, and ensuring governance and controls are in place to mitigate them. Insurers should conduct testing, model validation, audits, and plan meticulously, setting guardrails and ensuring that an owner is appointed and kept in the loop to manage at all times.
Powers that Be
The regulations in the insurance market will tighten with AI from ICO and GDPR requirements regarding data minimisation, transparency, and lawful processing. Alongside the FCA and PRA’s investigations into AI algorithms, model governance, and accountability, the upcoming AI regulatory frameworks will emphasise explainability, human oversight, and auditable controls. Companies that do not comply with these regulations will face fines, damage to their brand, and loss of customers.
Measuring Outcomes
To ensure your AI solution succeeds, regularly measure your outcomes to determine whether you are achieving your goals. Consider areas such as: minutes saved, customer satisfaction, whether you have increased revenue, reduced abandoned calls or queues, or if you have converted more quotes. It is helpful to establish specific measurable objectives from the outset so you know what you are aiming for.
Deconstructing AI
Reduce the fear of implementing an AI solution that is too costly or large to get approval for and to implement. Break it into smaller, manageable parts to achieve quick wins, but remember that preparation and the steps above should remain a priority for success. To simplify it, you can add a platform that integrates with the existing architecture so that if it doesn’t work out, it can be replaced. This approach is useful for traditional insurers beginning their digital transformation journey with ageing, complex, and siloed legacy technology.
AI for Success
Insurers investing in AI will personalise and enhance the customer experience, boosting process efficiencies and lowering costs. By laying a solid groundwork through alignment with your business goals and empowering people and processes, you will turn your AI project from potential into real performance. This approach will help ensure its success and the achievement of the targeted results.
