Nelson Castellanos, Chief Partnerships Officer – International at HDI Embedded.
Today, trying to convince anyone of the importance of artificial intelligence (AI) and machine learning (ML) is like preaching to the choir, especially with the rise of Open AI and new iterations of ChatGPT.
Unsurprisingly, many industries have jumped on the bandwagon including insurance, an industry known to have a slower rate of digital transformation. Organisations in this market have typically been more reluctant than most to shift away from their legacy tech stack. However, increased competition means companies are now innovating and embracing technologies like AI and Generative AI in particular.
AI is increasingly becoming invaluable for solutions like embedded insurance, driving automation to ensure much faster pay-outs to customers and reducing inaccuracies in those pay-outs. However—like with all technologies—the use of AI is coupled with some risks, including bias. That is why there have been some efforts to regulate the technology. Finding the right balance between not stifling innovation and encouraging bad actors is important.
Implementing efficient AI solutions in insurance
According to research done by Signicat, fraud prevention decision-makers across Europe are experiencing an uptick in AI-driven identity fraud and expect it to grow. However, AI can also be part of the solution.
We’re already seeing evidence that AI can reduce risk in the insurance space. AI-driven fraud detection systems can identify and mitigate fraudulent activity and ensure greater market resilience. This means minimising false claims, identity theft and exaggerated/staged accidents. It can do this by analysing vast amounts of data in real-time and then identifying and flagging anything that is out of the ordinary.
There are also benefits for embedded insurance more specifically. Embedded insurance is an innovative way to integrate insurance offerings into a third-party brand or platform. It already allows for dynamic pricing as insurance providers can tailor costs to the customer’s risk profile via data from open banking APIs. But AI’s ability to handle and process vast datasets effortlessly means embedded insurance providers can price policies even more accurately, develop more precise risk models, and streamline the assessment process.
The proliferation of data, facilitated by the growth and capabilities of AI will ultimately lead insurance providers to understand their customers better. Armed with richer information, insurers can refine pricing accuracy, tailoring premiums to granular risk profiles. This data-driven approach enables the creation of targeted insurance products, optimising coverage for individual needs. Additionally, open insurance practices, facilitated by data sharing among insurers, will help simplify the acquisition process, going even further in providing consumers with easily accessible and tailored insurance options.
With more power comes greater responsibility
There is a big caveat with AI around social responsibility. Successful AI applications are heavily dependent on training data – and if that data includes bias, it can lead to flawed outcomes. This proves crucial in cases like underwriting, where the legitimacy of claims can be judged by inherited human biases. For example, businesses seeking insurance could face inflated prices if an AI model assesses risk based on location or industry when the reality is AI should support more dynamic and tailored policy offerings.
This leads to a larger conversation of over-reliance on AI. Organisations need to find the right balance between human involvement and AI usage, particularly at these early adoption stages so the potential for AI errors and bias don’t go unchecked. It’s crucial that we see AI for what it is – a useful tool to support the human workforce, rather than one to replace it. The more this is done, the more trust in the models will increase.
The good news is that AI is already being hotly debated in the insurance world, with organisations and trade bodies creating guidelines and frameworks for how to use this technology safely. Meanwhile, Europe has now set a benchmark for AI laws in restricting high-risk applications, which may impact the use of AI in risk assessments and fraud detection. While guardrails around this technology are important, too much red tape could effectively mean significant regressions in the digital transformation of whole markets, and could become a key bottleneck to innovation.
Again, balance is crucial. Artificial intelligence (AI) is becoming a vital part of enhancing customer experience across the wider financial services space. So strictly regulating the use of AI due to the potential risks it poses could threaten the transformation that industries like insurance are finally embracing.
The future of AI in insurance
The future of insurance will see a transformative shift in the industry as insurtechs will expand into diverse, new markets. For instance, embedded insurance providers are seeing opportunities in areas like smart homes. Expansion into this market will allow insurers to integrate with emerging technologies, offering specialised coverage for smart home devices, security systems, and IoT-connected appliances.
By tailoring insurance products to specific needs, insurers gain even more data that can truly provide relevant and comprehensive coverage. This will attract a wider customer base and ensure insurers stay ahead in the rapidly evolving insurance landscape, driven in part by the introduction of AI.
