Let my AI email bot loose on a client and dodged a bullet

I was testing this new automated outreach tool for a real estate client in Austin last week. Set up a campaign to send follow-ups to leads, but forgot to check the filter for 'do not contact' list. Bot blasted 47 people who had explicitly opted out, and one of them was the CEO of a brokerage we were trying to partner with. Had to jump on a call and explain it was a glitch, managed to smooth it over by offering a free month of service. Has anyone else had an AI tool backfire like this and still salvage the relationship?

2 comments

2 Comments

gibson.sean1mo ago

Let my AI email bot loose..." Yeah that sounds about right. It's weird how we trust bots to do the stuff we'd never trust a human to do without double checking. Every time I see someone automating customer stuff I just assume there's going to be a "whoops" story behind it sooner or later.

-1

the_claire1mo ago

The one I remember most was a hotel chain that let an AI loose on their customer service chat. After about three weeks it had accidentally booked 200 people into rooms that didn't exist. The kicker was it kept apologizing and offering free upgrades to rooms that were already double booked. Management had to call every single guest individually to sort it out. The bot had no idea it was breaking anything because it didn't have any real understanding of what "available" meant. I figure that kind of chaos is coming for a lot of companies once they stop paying attention.