TDI pilot program testing AI to modernize policy reviews
'If we're going to regulate a high-tech insurance industry, we must have cutting-edge tools to keep up,' says Sullivan.
The days of adjusters laboriously poring over Excel spreadsheets to review insurance policies may one day be a memory at the Texas Department of Insurance (TDI). The department is currently testing the use of artificial intelligence (AI) to instantly scan and compare mountains of data. Insurance Commissioner Kent Sullivan sees AI as a “game-changer” in helping TDI modernize its practices to better protect consumers and ensure regulatory consistency.
In the pilot program, TDI will use AI to compare five years of auto insurance policy to see what language has been approved or denied. The program is currently in the data entry and concept testing phase. Phase two will use the system to review new policy filings.
Related: Leveraging AI for intelligent process automation
Drowning in paperwork
When companies want to sell a new insurance policy in Texas, they must file the policy language with TDI. Companies filed almost 30,000 property & casualty policies and related forms and almost 15,000 health policies with the agency last year — just one personal auto policy can be more than 40 pages long. Adding more complexity, today’s products are targeted to individual consumer needs with countless policy variations.
While TDI employees have reference charts and other tools to help with reviews, they also have to rely heavily on experience, carefully reading every policy and manually comparing provisions in similar policies as needed. Still, when language is unfair, they often need to dig through multiple databases to see if similar language has been approved in the past.
Related: Insurance game changers to watch in 2019
Faster process on the horizon
Because the AI review is nearly instantaneous, the pilot program promises dramatic improvement to the current process. When a policy is entered in the AI system, the system reads it and immediately sees whether there’s anything similar in the database. In addition, it can also look for required consumer protections and compliance with state law.
While it may seem that the AI process eliminates the role of the regulator, that’s not the case. Sullivan emphasizes that the decision-making process remains with the regulator — but now with a tool that offers faster and more comprehensive information on which to base decisions. “If we’re going to regulate a high-tech insurance industry, we must have cutting-edge tools to keep up,” says Sullivan.
Related: 4 things insurers should keep in mind when embracing new technologies