Artificial intelligence is becoming part of how companies present themselves to the world. It writes copy, answers questions, and talks to customers. We can argue about whether that’s good, bad, or inevitable, but from a branding perspective, the more immediate question is: what happens when it gets something wrong?
I was raised in a Catholic family. My parents eventually drifted away from the church, but my paternal grandparents were super Catholics, and when my folks would drop us off at their house so they could have a weekend away from us kids, my grandparents dutifully attempted to wash all our non-church-going sins away in one weekend. That meant church every morning and stories of sinners and sin.
One of my favorite stories my grandmother told me involved the host, the Eucharist, the little wafer we all receive at communion. My grandmother believed that you don’t chew the host. You let it dissolve so you get all the creamy holy goodness of this representation of the body of Jesus. She cautioned my brother and me never to let the host touch our teeth. That was akin to taking a bite out of Jesus. Like if Jesus showed up now and you walked up to him and took a chomp out of him. That’s what happens when you bite the host.
Even worse, when you do this, the wrath of God, focused on you for eating his only Son, is swift and deadly. You die. My grandmother had a list of names she’d rattle off, names of people who chewed the host and actually died, right there in their pew. The ol’ bite ’n’ die, which is very different from the bite-and-smile I used to do at commercial auditions, where you pretend to chomp into a burger and then flash the smile of a man who has finally found inner peace. Such is life.
Now, I believed my grandmother. Church terrified me: the high ceilings and stained glass windows, the smell of incense, and the image of a badly beaten man, head bleeding, his body straining against the nails in his hands and feet as his torso bowed away from the wooden cross. There was some kind of strange magic in that building, and I was caught in the middle of God and Satan fighting for my soul.
At my grandmother’s house was a book about the saints. Not the New Orleans kind. The kind who died terrible deaths and were visited by demons. There was one story about a saint who had a conversation with Satan, and the image that went along with the story chilled me to the bone. It was that image that always came to mind when Grandma talked about sin, sinners, and how all sinners go to hell.
Satan had horns and cloven hooves, a long serpentine pointed tail, and a ferocious mouth full of jagged teeth. For most of my young life, that was sin to me. That image of the devil perfectly represented the biblical warning: the wages of sin is death. And death, as far as I understood it, meant going to hell.
I don’t believe anymore that biting down on a communion wafer is going to result in my immediate death, and Satan has become considerably less vivid with age. But one part of my grandmother’s theology stuck with me: if you do something, you own what you did. There are consequences. Somebody has to answer for it.
Which brings me, rather unexpectedly, to artificial intelligence.
When the Sinner Is Your Brand
Artificial intelligence is becoming part of how companies present themselves to the world. It writes copy, answers questions, and talks to customers. We can argue about whether that’s good, bad, or inevitable, but from a branding perspective, the more immediate question is: what happens when it gets something wrong?
Not wrong as in a misplaced comma. Really wrong. It gives a customer bad information, invents something, or says something offensive. It does something nobody at the company would knowingly do and certainly wouldn’t want attached to the company’s name. Who owns that sin?
The temptation is to blame the technology. The AI made a mistake. The chatbot said it. The algorithm generated it. Somewhere in there is the comforting suggestion that the company was merely standing nearby when the whole unfortunate thing happened. Except the customer doesn’t experience it that way.
They didn’t choose your AI. You did. They didn’t decide where it would be used or how much authority it would have. Those were your decisions. The customer showed up and encountered whatever you decided would represent you. That makes the AI part of your brand experience.
And brands don’t get absolution simply because the thing that screwed up doesn’t have a soul. If your AI confuses, frustrates, or misleads a customer, they aren’t going home angry at artificial intelligence as a concept. They’re angry at you.
Maybe Grandma was onto something after all. Somebody still has to answer for the sin.
The Corporate Sin-Eater
I’ve written before about sin-eaters, people once paid to take on the sins of the dead by eating food placed on or near the body. The deceased got absolution. The sin-eater got dinner and, presumably, a fairly complicated relationship with God.
Companies have always had their own versions of sin-eaters. Something goes wrong, and there’s an immediate desire to find somewhere else to put the blame. A vendor screwed up, the system failed, or someone in another department made the decision. Now AI gives us something even better because it can actually perform the action without being capable of accepting responsibility for it.
The AI generated it. The algorithm decided it. The chatbot got it wrong. Three little sentences and the sin has been transferred, except it hasn’t. AI may be the perfect corporate sin-eater because it can swallow an extraordinary amount of blame without complaining. It won’t feel guilty, worry about its reputation, or lie awake at three in the morning wondering if it should have handled things differently.
The company gets to point at the machine, and the machine has no reason to point back. But the customer still knows who served the meal. You chose the technology, decided where to put it, and allowed it to speak for you. AI can do the work, and it can certainly make the mistake, but it can’t take the responsibility that comes with either.
We aren’t asking anyone to eat bread from the chest of the dead anymore. We’re asking a machine to eat the blame, but the sin still belongs to the brand.
Your Customer Doesn’t Care What Model You Used
When talking about AI, there’s a tendency to get very interested in the machinery. Which model are you using? How sophisticated is it? What can it do now that it couldn’t do six months ago? Those questions matter enormously to the people building and implementing the technology. Your customer, however, probably doesn’t care.
They care whether they got the answer they needed. They care whether the experience was easy or frustrating. They care whether they trust what your company just told them. The technology behind that interaction may be remarkable, but to the customer it is one more encounter with your brand.
We already understand this everywhere else. Nobody leaves a restaurant furious with the point-of-sale system. Nobody gets terrible service at a hotel and blames the software used to schedule the staff. The machinery disappears behind the experience, and AI doesn’t suddenly change that relationship. If anything, it makes understanding that relationship more important.
That matters because companies can get fascinated by what AI lets them do and forget to ask whether they should be doing it. Faster isn’t automatically better. Automated isn’t necessarily easier. Removing a human being from an interaction doesn’t improve the customer experience simply because you figured out how to remove one.
The question for a brand isn’t whether AI can do the job. It’s whether AI should be doing that particular job for your customer. Sometimes the answer will absolutely be yes. Sometimes putting a human being there will be slower, more expensive, and exactly what the experience requires.
That decision belongs to the brand. So does whatever happens next.
Efficiency Is Not an Identity
There’s a quieter sin, too. My grandmother would have called it a sin of omission, the thing you should have done and didn’t. Nobody ever dropped dead in their pew over one of those, but Grandma was very clear that they still counted. With AI, the sin of omission isn’t what the machine says wrong. It’s what you let it erase.
There’s nothing wrong with efficiency. If AI can take something tedious, expensive, or unnecessarily slow and make it better, use it. Nobody gets extra points for making customers wait because a human being was involved. The problem starts when efficiency becomes the strategy instead of something that supports it.
AI is increasingly available to everyone. Your competitors can use the same tools, chase the same efficiencies, and automate many of the same interactions. If everyone reaches for the same technology to solve the same problems in roughly the same way, the technology eventually stops being a differentiator. It becomes part of doing business.
That’s where a brand has to know itself. Your competitors can copy your tools. They can’t copy the reason people choose you. There may be interactions where speed is exactly that reason. There may be others where judgment, empathy, or another human being is what makes the experience worth coming back to. Knowing the difference requires more than knowing what the technology can do.
At ThoughtLab, we talk about creating a Category-of-one because differentiation doesn’t come from doing what everyone else is doing slightly better. It comes from understanding what is genuinely yours and building around it. AI can support that distinction, but it can also quietly erase it if every decision begins with what can be automated instead of what should be preserved.
The danger isn’t that AI will make every company bad. It’s that it could make every company familiar. Efficient, capable, polished, and increasingly difficult to tell apart. A brand should use AI to become more of what it already is, not more like everyone else using it.
Decide What Stays Human
The easiest question to ask about AI is what it can replace. We’ve been asking some version of that question since the technology became useful enough to make the answer interesting. Can it write this? Can it handle that conversation? Can it do something in thirty seconds that currently takes someone three hours?
Those are reasonable questions, but they start in the wrong place. Before a brand decides what AI can take over, it should understand what it can’t afford to lose.
That answer will be different for every company because what makes brands distinct varies. For one, it might be the way employees solve unusual customer problems. For another, it could be a particular point of view or a personal relationship customers have come to expect. Those things may be inefficient. That doesn’t make them expendable.
If you don’t understand what makes the experience distinctly yours, you won’t recognize when you’ve automated part of it away. It won’t announce itself. That’s the trouble with a sin of omission: nobody notices what’s missing until your customers do. You may save time, reduce costs, and increase output while slowly removing the very thing that gave people a reason to choose you.
None of this means keeping humans around for the romance of it. Some work should be automated. Some interactions can be improved enormously by AI. The point is to make those decisions deliberately rather than treating everything AI can do as something it should do.
Don’t ask only what AI can replace. Ask what your brand can’t afford to lose. Once you know that, deciding where the machines belong becomes a hell of a lot easier.
The Takeaway
I don’t know if there’s a hell for AI. I suspect not. Hell requires consequences, and consequences only matter if you have something to lose. AI doesn’t have a reputation to protect, a customer to disappoint, or a company it spent years building. We do.
That’s why the responsibility can’t follow the work when we hand it to a machine. AI can write the words, handle the interaction, and make the decision, but the brand still owns what happens afterward. No technological absolution waits for us when something goes wrong.
There’s something even more important at stake, though. AI isn’t only changing how brands work. It has the potential to change how much they resemble one another. When everyone has access to extraordinary technology, access itself stops being extraordinary. The advantage has to come from somewhere else.
That’s why becoming a Category-of-one matters more in an AI world, not less. You need to know what belongs uniquely to you before a machine can help you express it. You need to understand what customers value before you automate their experience. Most importantly, you need to know what you refuse to surrender simply because technology has made surrendering it possible. That’s the sin of omission no one will confess for you.
At ThoughtLab, Category-of-one has never meant being different for the sake of being different. It means discovering the space only your brand can occupy and having the courage to own it. AI can help you build there, but it cannot find that place for you. If you don’t know who you are before you hand the work over, the machine has nowhere unique to take you.
My grandmother believed sinners eventually had to answer for what they’d done. I’m considerably less certain about the mechanics than she was, and I remain hopeful that accidentally chewing a communion wafer isn’t an express ticket downstairs. But the basic idea still holds up surprisingly well.
Use the technology. Embrace what it makes possible. Let it make you better. Just don’t ask it to make you you.
That job still belongs to us.
