Can insurers make a profit with artificial intelligence?

Offering support across a variety of functions including customer service, claims processing, underwriting, and fraud detection, the capability of AI to analyze large datasets and process information will continue to revolutionize insurance.

Nirmal Ranganathan (pictured), vice president of engineering, AI, at Rackspace Technology, spoke with Insurance Business on how insurance companies can take advantage of AI to generate cost savings across their businesses.

The value of AI 

According to recent estimates by McKinsey, AI technologies could inject up to $1.1 trillion in annual value into the global insurance industry.

Of this sum, around $400 billion is projected to stem from enhancements in pricing, underwriting, and promotional technologies, while an additional $300 billion is expected to result from AI-driven advancements in customer service and personalized offerings.

What are the benefits of using AI?

Ranganathan noted that insurance companies often lack the capacity to conduct comprehensive audits on every claim they receive. Instead, they may opt to use sampling techniques to confirm the accuracy of their resolved claims.

Speaking on the benefits of leveraging AI for claims auditing, Ranganathan said: “Right now, auditing is a very people intensive task, but, if we introduce AI, you can go through all claims and provide a first pass accuracy check, and then a human can take a quick look at things to verify. If you’re able to catch theses errors that’s potentially millions in cost savings for a company.”

The wider industry recognizes these benefits, as evidenced by a recent survey from Gradient AI, which indicated that over 90% of insurers are planning to ramp up their investment in AI.

In a market characterized by nuance and complexity, Ranganathan highlighted AI’s potential to aid in policy comparisons, ensuring insurance professionals select the most suitable products for their clients.

Additionally, AI can streamline the processing of information and imagery in claim files, alleviating mundane tasks and enabling insurance professionals to concentrate on higher value problems.

“AI opens the possibilities for us to do other things – and that’s how over centuries society has grown. As things are simplified, we focused on the next challenging problems,” said Ranganathan.

While noting the many applications of AI in insurance, Ranganathan stressed that there will always be a place for human expertise.Top of Form

“There are a lot of scenarios where the human element is of value,” he said. “There is only so much an AI can reason. It can typically get you 90%-95% of the way but that 5% element usually requires a human.”

Third-party risks and bias

Despite the cost and time saving benefits that AI delivers, Ranganathan said that with more technology comes more responsibility.

He noted that when insurance companies opt to integrate third-party AI technology into their services, they must be aware of the providers’ terms of service.

“I think this is where a lot of people don’t have as much clarity,” he said. “Especially within the insurance industry. If you are an insurance provider and implement AI into any part of your product, you are still liable to explain what’s happening there, because that AI tool is being issued by you.”

“Pretty soon, every tool is going to have some form of AI. So, you need to decipher where exactly it’s being used and how it impacts your customers,” continued Ranganathan.

Given AI’s tendency to amplify human biases, Ranganathan also said it’s paramount that developers prioritize fairness in their algorithms to mitigate any potential adverse impacts.

“That’s where applying responsible AI practices can help in ensuring that there is proper transparency and accountability,” he said.