AI in insurance: The future is now
These use cases illustrate how implementing AI now benefits short-term carrier ROI while positioning those companies to be more sophisticated over time.
Since insurance companies started to adopt bots powered by Artificial Intelligence for customer service and claims about three years ago, the hype around AI in insurance has been building.
While cautious observers have been building the hype, smart carriers have found ways to recognize value by working with AI-focused startups.
There are two kinds of companies that are using AI to deliver value to insurers right now. First, there are startups that independently create more accurate data using AI. Re/insurers can then use high-quality data in their processes, instead of traditional, lower quality data. Second, there are startups that apply their AI technology directly to the data and processes of insurance companies.
Proof in the pudding
Here are a number successful use cases for AI in insurance involving Cape Analytics, a company that uses machine learning and computer vision to extract property data from aerial imagery:
Use Case No. 1: Inspection Costs: One carrier implemented Cape’s data as a way to reduce inspection costs. Using Cape’s AI-generated database, the carrier is able to sort properties based on the shape and condition of the roof, identifying properties for which an inspection is most likely to result in an actionable discovery.
The outcome: The carrier reduced inspection spend by 50% while maintaining underwriting quality.
Use Case No. 2: Identifying Erroneous Data. Another carrier used Cape to identify incorrect data received during application submission, which had led to systematic under-pricing for a segment of the book.
The outcome: The carrier recognized a 5-time return on its investment in Cape data, by pricing based on accurate property data.
Use Case No. 3: Workers Comp Loss. A commercial insurance carrier engaged Clara Analytics, which uses AI to reduce commercial insurance loss costs, to apply its AI-generated data to workers compensation.
The outcome: The carrier reduced loss costs by 11.6%. This benefit came from analyzing the specifics of the claim and deflecting claimants away from providers that were least likely to drive good outcomes for that kind of claim.
Use Case No. 4: Attorney Usage. Another commercial carrier hired Clara Analytics to reduce attorney involvement in workers compensation cases.
The outcome: The carrier reduced attorney involvement by 14% to 4%, through early identification of prospective attorney involvement, which the carrier’s existing models had not detected. In the second category, there are many companies delivering dramatic time and cost savings by applying AI to insurers’ internal data and processes, starting with transforming unstructured data to structured data, which can be used in decisions.
Three more successes
Use Case No. 5: Application Processing. A carrier wanted to expand its business into new markets but was constrained by how many applications its staff could process. The carrier hired Quantemplate to put in place AI-based processing technology.
The outcome: Quantemplate’s processes decreased the carrier’s headcount working on application processing by 67%, while simultaneously increasing its processing capacity by 50%. So, the carrier was able to penetrate a new market, while spending two-thirds less money to process the existing and new applications.
Use Case No. 6: Data Uniformity. QBE employees were spending 28.3 hours a week reviewing policies to ensure uniformity of language and endorsements. The carrier hired RiskGenius to implement its AI-based policy review technology.
The outcome: QBE employees now spend less than 13 hours on these tasks, a savings of over 50% of their time.
Use Case No. 7: Manual Data Entry. A carrier wanted to move away from manual data entry, so engaged HyperScience to apply its AI-based technology.
The outcome: The carrier decreased its error rate by 67%, while also reducing the time to process data by 5X. As a result, the carrier was able to reduce the number of employees allocated to this task by 80%, while delivering substantially more accurate data.
Connecting the dots
The common themes across these case studies are that carriers are saving time and money while improving data quality. These improvements should lead to better underwriting decisions, higher margins, and the ability to serve new markets and new customers.
Carriers that make the decision to implement AI now will not only benefit from short-term ROI, but also position their organizations to be ready to implement more sophisticated AI, over time.
If insurers want to run their operations faster, better and cheaper, now is the time to embrace AI.
Martha Notaras (martha.notaras@axaxl.com) is a Partner at venture capital firm XL Innovate, where she invests in new business models as well as data and analytics startups creating insurance-specific solutions. Martha has over 20 years of investing and corporate development experience in technology, information, and financial services companies, and has served as board director for over a dozen portfolio companies. Her current investments include Pillar Technology, GeoQuant, Notion, and Cape Analytics.
These opinions are the author’s own.
See also: Top insurance technology issues nagging at industry leaders