Charles Taylor on challenges in implementing generative AI

At Charles Taylor InsureTech, we enable insurers, brokers, and MGAs to progressively and strategically integrate generative AI into their operations. 

The specific return on investment varies according to individual use cases, such as automating tasks or generating content. 

However, our clients consistently seek — and achieve — operational benefits by investing in tools to optimise processes, simplify channels, and enhance experiences both for end customers and platform operators.

All these initiatives have a positive impact on the bottom line, and we’ve seen appreciable cost savings thanks to reduced response times, automation of repetitive tasks, and more informed and effective decision-making. 

As for opportunities, there are many exciting developments across the insurance value chain, from distribution and applications to risk assessment and claims processing.

Collectively, these enhancements have led to greater operational efficiency and significantly improved customer satisfaction. 

Viewed globally and industry-wide, Satisa reports that 41% of organisations in 2024 achieved a return on investment of between 11% and 33% for generative AI initiatives.

Challenges

One significant challenge in implementing generative AI with our clients is ensuring AI models effectively interact with existing systems and provide accurate responses across diverse queries and scenarios. 

To address this, we’ve developed a platform that allows dynamic creation and modification of specialised APIs through low-code tools. 

These APIs guide our AI agents, enabling them to access precisely filtered and contextualised data, improving both accuracy and response times. 

Having specialised and structured APIs is crucial, particularly in real-time interactions such as voice or chat, where efficiency and rapid response are essential for user satisfaction and overall system performance.

There is a steep learning curve with every new or emerging technology, and adjustments and guidelines will be essential as generative AI use cases develop. 

One of the main challenges is to create a framework that ensures continuous legal and regulatory control without impeding adaptability and flexibility. 

Because generative AI is such a dynamic and fast-evolving technology, it is important to match innovation and speed with control and trust. 

We must remain alert to inaccuracies creeping in from flawed data, and maintain security and privacy. Platforms that support dynamic, low-code API modification will be vital for quickly integrating new functionalities and adapting to market forces and regulatory demands.

As for opportunities, there are many exciting developments across the insurance value chain, from distribution and applications to risk assessment and claims processing. 

Moreover, new use cases are emerging all the time because generative AI is such a versatile and creative technology. There is significant potential to enhance omnichannel capabilities, enable agents to model workflows on the fly, and improve customer experience and hyper-personalisation. 

Using advanced AI agents and specialised APIs, we can simplify complex processes and minimise manual intervention; offer highly personalised experiences; and boost operational efficiency across a wide range of tasks such as predictive modelling and fraud detection.