For this Regulatory Brief, AAIS sat down with George Bradner, Assistant Deputy Commissioner of the Connecticut Insurance Department, to discuss the use of modern technologies to streamline processes for more efficient filings moving into 2023.
"In our department, as in many departments around the country, we have extensive institutional knowledge," shared Bradner. "We have people with 20-40 years of insurance experience with forms and knowledge in regulation. When people retire, knowledge leaves. We want to create a process that uses AI (Artificial Intelligence) and machine learning to take that institutional knowledge and replicate it so that if someone leaves that knowledge is still here, and we can use it to also train new people when they come on."
Bradner also discussed a current project Connecticut DOI is working on with AAIS to utilize AAIS Machine Learning powered by KIRA. According to Bradner, the technology allows you to train the system and it will learn based on how it is trained. When an examiner is reviewing a filing, for instance, "the system can build out the rules and train KIRA on what is acceptable and what is not. Once that is done a hand full of times, the system will be able to analyze a policy when it comes in,” accurate over 95% of the time. It speeds up the review process and provides consistency in the review."
Bradner also updated us on a very successful pilot looking at claims made. "One of the things we learned is that we did not have a lot of great compliance with our claims made regulation. KIRA pointed that out.” Looking ahead, "we are identifying the rules we need in order to begin training KIRA to review automobile forms filings."
Bradner feels the application of AI and Machine Learning is easily transferable to other states. "If you have a regulation that is very detailed or checklists that help an examiner in their review process...you can program that into the system. Ultimately I hope to put that out front so when a company makes a filing with us it has to go through the system before it's accepted within SERFF and transmitted within the department."
To view the full interview with George Bradner, please click the video above.