Tokio Marine has teamed up with AI powered claims technology provider Tractable to accelerate its claims processing. The partnership enables the insurer to use computer vision to assess vehicle damage in near real-time and estimate the extent of repairs needed. Tractable is also operational at other insurance giants such as MS&AD (fifth largest P&C insurer) to help accelerate servicing across thousands of auto claims each year.
Tokio Marine has of late, solidified its partnership with catastrophe risk modeler RMS to scale risk insights on perils and markets at enterprise level. And in its latest announcement last week, the insurer announced a partnership with leading insurtech Lemonade to enhance its operations. It hopes to create a range of new products and services through this partnership.
So, what’s the big deal about insurtech innovation, especially in intelligent tech such as AI that is seeing a surge in partnerships?
In the case of Tractable, MS&AD believes it will help adjusters save an average of 360,000 hours of time each year. Tractable’s AI has been trained on photos and human repair decisions spanning millions of historical accidents and can assess damage to PVs in Europe, Asia and Americas.
Spread over more than 25 banking use cases, a study on “Meeting The AI Challenge” reveals AI technologies can help boost revenues through increased personalization of services, reduce costs through efficiencies generated by automation, reduce error rates and improve utilization while tapping unrealized opportunities.
The number of times AI related terms appeared in investor annual reports has increased from two citations in 2015 to 116 last year. The number of AI-related patents filed by insurers between 2010 and 2019 increased by more than a factor of 50, with 10 US insurers accounting for 80% of activity. In China, less than 5 insurers accounted for 85% of AI-related patents filed.
Conventional AI such as Generalized Linear Models (GLM) have become commonplace for risk assessment and prediction models. Of late, enthusiasm for other techniques such as deep learning and reinforcement learning have been on the rise, seeing an increase in pilots conducted. Early adopters see benefits in areas such as faster claims settlement, more targeted cross- and up-selling, improved fraud detection and better risk scoring.
From Github, the latest activities in intelligent tech across few verticals are predominantly in traditional areas of analytics, followed by AI. Newer advances in machine learning (ML), deep learning are still fewer in number.
An example of a featured organization with a mature AI setup is CCC Information Services that provides tech solutions and services for automotive, insurance and collision-repair. It helps insurers to improve and simplify claims processes through telematics and AI. They have 75 different AI models running with 250 in the pipeline, built by a team with dozens of doctorates in AI, ML and physics. Setting up an organization that can deliver AI at scale is not easily replicable.
Additionally, there are deterrents to deploy intelligent tech at enterprise scale, such as:
- Collecting and curating relevant data by addressing privacy regulations and incentives, fragmented processes.
- Assessing and processing relevant input information provided by AI.
- Trusting intelligent automation when there are “obvious” misses.
- Usefulness is undermined when data does not rapidly translate into actionable insights.
Partnering with insurtechs has been a convenient path for incumbents to break through some of these hurdles. An example in the life space is PartnerRe, which formed a partnership with Ladder. Founded in 2015, Ladder streamlines life insurance underwriting using ML techniques and offers affordable policies. PartnerRe President Chris Shanahan admits, “We could not be more excited for the opportunity to work with the team at Ladder and their existing partners. The progress that they have made together is impressive and we’re eager to contribute and support Ladder’s future growth.”
You get 3 free articles on Daily Fintech. After that you will need to become
a member for just US$143 a year (= $0.39 per day) and get all our fresh
content and our archives and participate in our forum.