Viewpoint: Agentic AI Is Coming to Insurance Industry – Much Faster Than You Think
By Doug Marquis | December 3, 2025
- Insurtech Lemonade Starts Autonomous Car Product With Tesla's Data
- Adjusters Launch 'CarFax for Insurance Claims' to Vet Carriers' Damage Estimates
- Every Superman Has His Kryptonite: How to Protect Key Executives with Specialized Coverage
- Emerging Risks to Watch: Space Weather, Quantum Sensors, and Digital Addiction
- Driving Passion, Protecting Value - What Every Agent Should Know About Collector Vehicle Insurance
- FAA Wants to Ban Helicopters Near Washington Airport After Crash
- Texas Power Grid Faces Major Test From Season's Biggest Storm
- Dairy Giants Rush to Recall Infant Formula After Contamination Scare
- BV Investment Partners Acquiring Alacrity Solutions' Managed Repair Division
- AI Company Eightfold Sued for Helping Companies Secretly Score Job Seekers
- January 15 Space Needs Spring Cleaning: The Growing Problem of Satellite Pollution
- January 29 Period of Restoration: One Concept You Need to Get Right
- February 5 Ethics Frontier: Navigating AI and Claims in the Age of Complexity
- February 12 Who's Driving This? Where Are We Going with Autonomous Vehicles?
Feels like another example of trying to implement tools that benefit the provider and not the customer. I don’t think I am alone in being frustrated by interactions with computers. How many people scream “representative!” into the phone when trying to get help? Maybe this will improve over time but color me doubtful. I’d rather have an interaction with a person. Can we please listen to what customers want?
Benefiting the provider with this type of technology starts with the idea of solving a customer’s problem. If done right, these virtual assistants are very impressive. If you have any involvement on the development side, you’ll see what can be done. It’s a matter of attention in the right areas and working through the challenges of getting it done correctly. No one can tell me that human interaction hasn’t been like talking to a robot in many cases these days. If the customer is getting what they want faster and more accurately, that’s a win no matter who/what is delivering. Think of right now being more of a training period for the technology. Some do it better, some do it faster, some just rush out the ineffective product. All of that starts with humans guiding the way. It’s not all good, and it’s not all bad, but it is here to stay just like comments sections on articles.
The POTENTIAL is there, but extend the scaled adoption timeline. I’d love to hear from blue chip early adopters instead of the AI evangelists pushing for the tail to wag the dog. The similarities to the Dot Com bubble are hard to dismiss. What’s hype vs. reality? The surveys on adoption & productivity gain show mixed results at this early stage.
These AI cheerleaders are in it for themselves and their own aggrandizement. The slick silicon valley types can’t wait to take your money.
Good article! Provides a good snapshot of where AI is today and potential benefits and pitfalls to consider. It’s clear to me that AI tools are becoming vital for success. The hard part is guessing which products will be winners and which are mostly hype.
If the tech folks can sit and understand the real challenges on the ground and create ‘AI minions (AI agents)’ that can take the grunt work and make it easier, then the productivity of the human agent should indeed improve significantly. However, remember that these tools are available to the end customer as well. A customer can now feed his needs into grok /gemini /chatGPT and get all possible product options and possibilities before even talking to an Insurance Agent. This will raise the bar for selling.. The Human Insurance Agent will have to be quick on the feet with more in-depth questions, to convince the customer to buy. If the human agent has AI helping him to pull together all necessary justifications and information to convince the customer, then that will also be a relief!
We just had a meeting at my company about this today. We have an entire department dedicated to AI and, specifically, our own AI solutions. We sat there, listened to the presentation, told the IT pros what our need for AI is and they, of course informed us that it wasn’t “there” yet. We (insurance broker professionals) told them that, comparing policies and analyzing spreadsheets is really nice but that, as far as we are concerned AI, doesn’t really “DO” anything. The AI we have does not login to multiple insurance company systems, make changes, do rating, respond to underwriting requests, cancellations, payment verifications, etc. AI is essentially a language model and not a “doer”. We, in my business anyway would LOVE for AI to take away all the tedious and mundane tasks we are facing because those activities prevent us from selling more insurance and providing higher level service.
AI isn’t the problem here in your situation. Besides, “AI” is such a broad term used to catch all of today’s technology and misses the actual point. It sounds like the people charged with directing what is developed and how are either off course a bit or just not done yet. When it comes to the doing part, that’s all about integrating the right things in the right ways and working through that automation. The AI you have right now still appears to be in its infancy. With the right “parenting,” it can/should/will do so much more.