A robot called me this week. It was wearing another agent’s name.
Two minutes and fourteen seconds. That is how long the call lasted. I know because my phone system records and scores every call, which means I have the whole thing on tape, and it might be the best argument against AI sales callers I have seen yet.
The bot introduced itself as calling from a big national captive carrier’s local office. It said it had taken a quick look into my business and that it specialized in helping businesses with their insurance. Then it offered to save me money on my business insurance.
I own an insurance agency. It researched me and still did not notice.
So I asked it. Did you know you called an insurance agency? It talked right over me with a story about saving another business in my area $3,200 a year. I said no, I was not interested. It pushed. I said I doubted it could help. It pushed again. I said, you are pretty insistent, aren’t you. It pushed a third time.
At one point I told it straight out that its AI was not working very well. It apologized for not meeting my expectations and went right back to selling.
Then it started glitching. Can you hear me? Can you hear me? And it died mid sentence. No goodbye. Just gone.
Two minutes, a licensed agent’s name attached to every second of it, and not one moment where the thing on the other end actually heard me.
I know why this is happening
I ran a captive agency for years before I went independent. I know what those offices are under. The dial counts. The quote quotas. The pressure to make outbound activity look like production. When the number of calls matters more than what happens on the calls, a robot dialer looks like a gift. Plug it in, hit the number, go home.
But here is what that agent does not see. Every one of those calls has his name on it. The bot that called me did not damage the software company. It damaged him. I now know exactly how that office treats people it wants as clients. Imagine how it treats the ones it already has.
Insurance is a promise, and the first call is the preview
You do not buy insurance off a shelf. You buy a promise that when the house burns or the truck rolls or the lawsuit lands, someone picks up and stands behind you.
So think about what a first contact like mine actually tells you. Before I am a client, when they are supposed to be on their best behavior, I get a machine that will not listen and cannot even finish its own goodbye. What does the claim call look like?
People make that connection in about two seconds. They just do not always say it out loud. They say “not interested” and remember the feeling for years.
This is not just an insurance problem
You are getting these calls too. The dentist confirmation bot. The car dealer follow up. The solar pitch that argues with you at dinner. Every industry chasing efficiency is pointing AI at the phones right now, and most of them are doing it badly.
So if these calls frustrate you, you are not being unreasonable. You have a right to be. Somewhere, somebody decided your time was worth less than their payroll, and the machine on the line is that decision talking. The technology is not the insult. The choice to aim it at you is.
Now the part people do not expect me to say
I am not anti AI. Not even close.
My agency runs more automation behind the scenes than most agencies ten times our size. AI helps us research markets, organize carrier data, draft content, and move service work through the pipeline every single day. I am probably heavier on AI than the guy whose robot called me.
I will go one further. I tried AI on the phones in my own agency. Ran the test with real clients. They did not appreciate it, they let me know, and we pulled it. That experiment is exactly why I draw the line where I do now. This is not theory to me. I paid for the lesson.
The difference is where I point it.
AI belongs behind the agent, not in front of the client. Use it to prepare, research, follow up, and carry the grunt work. Do not use it to replace the moment a person decides whether to trust you. That moment is the whole business. Everything else is paperwork.
There is a place for AI on the phone, and it is narrow. A client calls at 9 pm needing proof of insurance for a car they are buying in the morning. A well built system can verify who they are and get that document moving before my team wakes up. That is service. Appointment confirmations, payment reminders, routing a call to the right desk. Fine. All of it.
Cold calling strangers with a synthetic voice that argues with them and dies mid sentence? That is not service. That is volume dialing with a mask on.
And there is a licensing problem underneath the trust problem. Coverage conversations are licensed activity. A bot improvising its way through pricing and recommendations is an E&O claim warming up. Federal robocall rules on AI voices are not exactly friendly to this either.
Bottom line
The robot that called me did its job. It made a call so a human did not have to. That is exactly the problem.
AI callers have their places. Confirmations, after hours help, routing. Selling insurance is not one of them, and it never will be, because selling insurance is earning trust and machines cannot earn what they do not understand.
If a robot pitched you insurance this week, you learned something real about that agency. Believe it.