Innovating the insurance sector with AI and automation

May 11, 2022


In brief

  • AI and automation are transforming the insurance sector by enabling companies to automate manual processes, analyze large volumes of data, and deliver personalized products and services to customers. 
  • Use cases include fraud detection, underwriting, claims processing, and customer service. 
  • Companies can start innovating by defining their goals, assessing their data and technology infrastructure, identifying use cases, and partnering with the right vendors and experts. 


Times are changing. The insurance sector is expanding its digital footprint. Industry leaders can’t escape the fact that now there’s a much bigger need for analytics and a far greater need for automation. But how do we use technology to mitigate exposure to fraud and where do we need to be innovating?


The right people for innovation


A lot of people wonder how to be innovative in fraud detection. When asked, many practitioners will cite that they have innovative models or technology, but much of this is very similar to other solutions or even the same solutions rebadged. To be really different, it’s important to hire different types of people (contrast that to the cybersecurity industry where it’s important to hire more people). Insurers have long employed people from detective backgrounds and there are many employees who are very skilled claims analysts; and they’re very competent. But now progress is all about numbers and analytics — a significant change for the industry and a change which requires a different skill set in workers.

Engaging with insurance organizations on a daily occurrence, it’s interesting to see that these changes have started already. Until two or three years ago, it was rare to meet someone with a data science title or an AI-focused role. Now though, there is a significant increase in the number of roles in AI and data science — a sign of the changes that are needed.


Why isn’t the use of AI and automation ubiquitous?


If you talk to insurance fraud teams on an operational level, more than 50% of people don’t understand what AI is and what it can actually do. There’s also some resistance to bringing in AI and automation, because you don’t need the same resource models you already have. Technology can easily uncover what you’re struggling to find with 50 people, and the tech can automate this detection without the use of people. A worrying point for those 50 people working in fraud detection? Not really. It shouldn’t be a cause for concern for the employees, as people can’t be removed from the operation — technology can’t tell someone they’re not having a claim paid for example. Rather than threatening jobs, AI tools and automation can improve and simplify the work of employees, giving them more time to concentrate on more stimulating tasks.

When it comes to innovation, there’s a major opportunity to improve the life cycle for insurance fraud employees. A lot of their work is repeated hourly, performing the same procedures every day. The vast majority of these tasks could be automated, allowing the people to perform in a more effective and fulfilling way. Automation can free up employees from the mind-numbing tasks (also performing them at a significantly higher speed) to help increase work efficiencies and raise satisfaction levels.


How to use AI and automation effectively


AI is fundamental now, because the insurance sector is pushing it forward at all levels. At the top, board members understand there is no escaping AI, so they’re asking for AI strategies and assessments of its current use. What is needed is more education — an improvement of the journey for people in terms of AI, instead of labeling what is already in use as ‘innovation’. As an example, many insurers now are tackling fraud a bit late — when people make a claim, rather than before providing a quote online. Some insurers receive requests for hundreds of thousands of quotes a day — the smart ones are using AI and automation to sift through these quotes automatically and separate good from bad in real time. But the potential goes beyond this one example; we should look at the whole process.

There is still a great need for better use of analytics and much more automation — two things that would allow people skills to be used in a very different way. The fraud (and broader financial crime) teams of the future will unpick the current fraud model and start anew with questions like ‘where do we need people? And what are they actually going to do?’ because the technology we have can make life better. Ultimately, insurers are there to do good things for customers — making the right changes now will improve the client journey and the claims process.

There’s a lot to do in the sector, but there’s a lot of good going on now too. Some companies are quite ahead, but a large number of insurers are stuck in the past — for them to survive, changes will need to come from the top as the operational people are pushing uphill. For some, there is resistance to change, but the most successful insurers tackling fraud will be (and currently are) the ones using AI and automation.

Luxoft, DXC’s Analytics and Engineering business, is the preferred partner for financial institutions where undertaking strategic transformation, and is providing fresh insight into insurance sector analytics. To learn more, contact:


Download pdf version