AI Is Here. Should You Actually Use It?

I came across something interesting recently.

A professor at Harvard Business School produced a case study. The case is called "Generative AI in Marketing."

It tells four short stories. Four stories from completely different industries.

Retail, FMCG, luxury, and B2B industrial technology.

Nothing unusual there, right? Every industry is talking about AI.

But the question this professor really wants to ask comes down to one thing:

Does the value AI creates for you actually outweigh the value it could destroy?

That question, I think, gets to the root of it.

What These Four Stories Are Actually Doing

These four scenarios correspond to four ways AI is used in marketing:

Personalized recommendations. AI helps you push different content to each user.

Synthetic research. AI helps you "create" respondents for user research.

Creative design. AI helps you produce posters, copy, and visual concepts.

Mass content production. AI helps you churn out hundreds of articles and thousands of creative assets in a single day.

Each one sounds tempting. High efficiency, fast turnaround, cost savings.

But think about it — these four things are precisely the four most sensitive wires between a company and its consumers.

Personalized recommendations, done poorly, turn creepy. Users feel like you're spying on them.

Synthetic research, done poorly, gives you conclusions that AI fabricated — bearing no connection to real users.

Creative design, done poorly, makes your brand start looking like everyone else's. Because everyone is using the same models, generating the same styles.

Mass content production, done poorly, makes your blog read like a robot factory.

That's what this case study is really about.

Value Creation and Value Destruction — Two Sides of the Same Coin

What does value creation mean?

AI saves you money, saves you headcount, saves you time. A week's work done in what used to take a month. That's value creation. Nobody denies it.

But here's the thing. While creating value, AI is quietly doing something else at the same time.

It's betting your brand.

Reputational damage. An AI customer service agent says the wrong thing, and it spreads across the internet in two hours.

Brand dilution. When your visuals, your tone of voice, your personality all become model-generated output, what's left that's actually "you"?

Legal risk. AI-generated content — who owns the copyright? If it "learned" from someone else's work to produce your images, who's liable for infringement?

You see, the efficiency side is visible. Money saved, reports looking good.

The risk side is invisible. Until one day it suddenly blows up.

Every gift from fate has already been priced in secret.

That line fits perfectly here.

So What Is This Case Actually Teaching?

It's teaching one thing: how to think clearly about whether to use AI, and where.

The case designer isn't trying to tell you "AI is great" or "AI is dangerous."

She wants you to work through these four scenarios yourself, step by step, and derive your own judgment framework.

Which scenarios: go ahead, let AI run free.

Which scenarios: tread carefully.

Which scenarios: don't even touch.

This framework isn't handed to you. It grows from your own thinking.

That's probably the mark of a great case study — it doesn't give you answers. It gives you a yardstick.

A Personal Reflection

Honestly, after engaging with the thinking behind this case study, I felt a strong sense of resonance.

Because so many companies are talking about AI right now. The conversations are lively.

But the vast majority of them stall at two extremes.

One extreme: "This is amazing! Get on it now! You'll fall behind if you don't!"

The other: "No, no, what if something goes wrong? Let's wait."

Both are wrong.

If you only see the efficiency and charge in, sooner or later you'll pay the price in brand and legal damage.

If you only see the risk and freeze, sooner or later you'll be left behind by the people who "thought it through before moving."

The hard part is which stage to act in, how far to go, and where to draw the line. Whether to do it at all is the simplest question.

The value of this case study is that it forces you to put these questions on the table and work through them one by one.

Not to find a universal answer. But so that before you press that button, you know exactly what you're pressing.

I don't know which stage your company is at right now. But I suggest — while there's still time — you think through these four questions first.

Efficiency is visible. Brand is invisible. And invisible things are often the most expensive.