AI and Machine Learning are driving benefits and reducing headcount in the insurance sector

A survey of insurers by Rackspace has found that the implementation of Artificial Intelligence (AI) and machine learning (ML) technologies are driving benefits and enabling the reduction of headcounts.

The survey found over the past 12 months 62% of insurers had cut staff numbers due to the implementation of AI and ML. They found that the new technology enabled low-level analyst work to be completed by AI and ML. 52% of respondents said they had already realised substantial benefits from AI/ML, with a further 23% seeing modest benefits. 25% of respondents said it was too early to tell.

The benefits insurers listed from implementing AI/ML were

• 81% risk reduction, increased understanding of business/customers

• 79% increased sales

• 77% personalised marketing

• 75% increased productivity

• 73% increased revenue streams, operation cost reduction

• 69% improved customer satisfaction

• 67% faster time to profitability, reduced cost of new product development, ability to hire/recruit new talent

• 65% increased innovation

There are still some issues with AI, with 42% only ‘slightly trusting’ AI/ML results compared to 28% ‘strongly trusting’ results.

Talent and skill shortages were seen by 67% as the greatest challenge to further adoption of the technology; however 90% of insurers had grown their AI and ML workforce in the past 12 months.

Although this survey lumps AI and ML together there are some fundamental differences. Machine learning can be disconnected from the large volumes of training data that are used in generative AI such as Chat GPT. Although that may sound like a disadvantage, narrower pools of training data can make results from machine learning applications much more accurate for highly specific tasks. It also means that data does not have to be shared with generative AI engines – its can remain in confidential silos within the business.

ChatGPT recently put forward it’s arguments for how it can positively impact the insurance industry. While there are still plenty of cons to using ChatGPT right now, ranging from inbuilt biases and prejudices to its failure to comprehend nuance such as sarcasm, some insurance executives believe the underlying technology could be used as a starting point to build on and to jumpstart innovations in the sector.

At Quality Product Research Limited we agree that there are some great opportunities for implementation of AI and ML initiatives. With more than 15 million quotes and over 1.25million data points of data in our research databases we are well positioned to employ these tools to greater effect over the coming years.