Man vs Machine – Can AI be Trusted for Auto Claims?

Artificial intelligence (AI) is a hot topic in many industries today. The promise of AI is that it can automate many complex and redundant human tasks, allowing companies to free up human capital for more productive uses and reduce costs. Many industries from automobiles (Tesla) to entertainment (Netflix) to financial services (Betterment) are successfully leveraging AI in commercial products today. In the insurance sector, several companies have started to use AI for sales automation, underwriting, and even claims processing. But can AI be trusted for such mission critical customer-touching functions? In a word – absolutely!

What makes the insurance industry so adaptable to AI is its deep dependency on data for core business processes. Carriers have long used data to identify and target new customers, quantify customer risks to underwrite new policies, and assess and settle claims. These processes also produce vast streams of data that can be analyzed to identify patterns and make future predictions – perfect candidates for AI automation. Several early initiatives are yielding incredible results.

In the claims area specifically, a number of firms have built AI technology that provides instant estimates of accident damage via uploaded photos from a mobile app, with accuracy often exceeding human-generated estimates. Carriers like Ping An (China) are using the technology today reduce claims leakage by hundreds of millions of dollars a year. And my own company, Claim Genius, is working with major US and Indian clients to provide the same.

Leading Australian P&C carrier Suncorp is using technology from IBM that automatically analyzes claims documentation and determines liability with over 90% accuracy, greatly reducing the need for human review. In the US, home insurer Lemonade is using AI chatbot technology to process claims in less than a minute with over 95% customer satisfaction levels. And Allstate’s new AI and NLP-based call center technology, called “Amelia”, has reduced call times for claims inquiries by over 30 seconds.

Like any new technology, AI has its limitations – AI claims apps have not yet mastered certain damage types like flooding, and AI-based customer service agents can often get confused by heavy accents, for instance – but the bottom line is that AI is providing carriers with new tools to accelerate claims settlements, dramatically reduce costs and fraud, and increase customer retention.

And this is just the start. As AI models are “fed” with more and new types claims data such as telematics crash data, GPS tracking data, medical and legal claims data, and the like – carriers will be positioned to greatly expand the types and accuracy of claims decisions that can be completely managed via AI. According to McKinsey, over 90% of claims will be completely “touchless” by 2030, driven almost entirely by advancements in AI.

The AI revolution in insurance is well underway. Carriers that have started to deploy AI solutions have already proven impressive results that are positively impacting their bottom lines and driving competitive advantage. It is imperative for all carriers to begin to leverage AI to position themselves optimally for future growth and success.