GenAI Won’t Replace Competitive Intelligence—But It Will Reveal Whether You Learned the Trade

In my award-winning, international best-selling book, The UnConventional Strategist, I wrote that GenAI is impressive. It can write essays, summarize meetings, recommend dinner spots, and explain Prussian military strategy like a caffeinated librarian.

But judgment?

Yeah … not quite there yet.

The same applies to Competitive Intelligence.

GenAI can perform Five Forces and Four Corners analyses, build competitor profiles, analyze financials, develop scenarios, generate hypotheses, and even play your competitor in a war game.

What it can't give you is years of experience doing Competitive Intelligence.

It can't give you judgment.

And therein lies both the extraordinary opportunity—and the danger.

Knowledge Is No Longer the Constraint

For most of my career, finding information was hard. We searched databases, called people, attended conferences, interviewed salespeople, read annual reports, tracked executives, and hunted for that one obscure piece of information that might change our analysis.

Then came the Internet.

Then search engines.

Now GenAI.

We've gone from information scarcity to information abundance.

GenAI can research a competitor, analyze its finances and product portfolio, identify strategic signals, compare capabilities, and generate hypotheses about what it might do next—in minutes.

GenAI has made knowledge abundant.

It has not made expertise abundant.

That's the difference.

The Intern Who Never Sleeps

I tell people to think of GenAI as a brilliant young intern.

They're incredibly smart, extraordinarily well-read, work faster than anyone you've ever hired—and have almost no practical experience.

Would you hand them the task of coming up with your company's competitive strategy on Monday afternoon?

Probably not.

You'd give them assignments. Review their work. Challenge their assumptions. Ask where the information came from. Point out what they missed.

That's how I approach GenAI.

Except my intern works 24 hours a day, reads extraordinarily fast, never complains, and doesn't mind when I ask it to redo an analysis six times.

Pretty good intern. But still an intern.

I don't just ask GenAI for answers. I manage it. I challenge its conclusions. I ask what's missing, what contradicts its hypothesis, and what would have to be true for it to be wrong.

That requires knowing the trade yourself.

The 85% Problem

I'm less worried when GenAI is completely wrong than when it's almost right.

Suppose it analyzes a competitor and correctly identifies its products, executives, financial performance, market position, strengths, weaknesses, and strategy.

Then it concludes the competitor won't enter an adjacent market because it lacks the capability.

Except… they just hired the executive who built that business at another company.

Or established a partnership that gives them the missing capability.

Or their CEO quietly signaled that their current market can't support their growth ambitions.

The analysis may be 85—or even 95—percent correct.

But the part it got wrong is the part that matters.

That's the 85% Problem.

The dangerous GenAI output isn't nonsense.

It's an analysis that's beautifully written, logically constructed, professionally presented—and almost right.

That's where experience matters.

Knowledge Parity Changes the Game

Something else is happening: knowledge is becoming commoditized.

Two competitors can use similar GenAI systems, access similar public information, ask similar questions, and get sophisticated answers within seconds.

So where's the advantage? Probably not in the answer.

I call this knowledge parity.

When everyone has access to similar information and GenAI capabilities, the advantage moves elsewhere:

The questions we ask.

The methodologies we use.

The signals we connect.

How quickly we recognize change.

Our judgment.

And ultimately—our actions.

The intelligence question is shifting from:

What do we know that they don't?

to:

What can we understand from what everyone knows that others haven't recognized yet?

That's where Competitive Intelligence gets interesting again.

Stop Looking for the Magic Prompt

Organizations are spending far too much time looking for the perfect prompt.

Guess what? It doesn't exist. A prompt is an instruction.

A methodology is a system.

I don't want GenAI to produce one impressive competitor analysis. I want a repeatable intelligence capability.

I want it examining strategy, products, finances, capabilities, markets, and behavior. I want it challenging assumptions, searching for contradictory evidence, developing competing hypotheses, identifying weak signals, and recognizing when something changes.

That's what led me toward a structured approach like the CGEN methodology from the American Institute of Artificial Intelligence (AIAI) and the Society for AI in CI (SAICI).

SAICI’s CGEN doesn't treat GenAI as an electronic oracle.

It creates a repeatable methodology around how intelligence is collected, analyzed, challenged, validated, and turned into action.

There's a huge difference between asking:

“Analyze Competitor X.”

Versus building a system that continuously examines Competitor X through strategy, product, financial, market, and behavioral lenses; identifies changes; compares those changes against previous expectations; generates competing explanations; identifies potential inflection points; and then tells the human practitioner:

“Something changed. You should look here.”

The first gives you an answer.

The second gives you a capability.

That's the fundamental idea behind SAICI’s CGEN.

GenAI doesn't replace the methodology. The methodology disciplines the GenAI.

Demonstrate Competitive Advantage. Don't Demonstrate AI.

We've all seen the GenAI demo. Someone types a prompt. Words appear. People nod. Someone says, “Wow.” Then everyone goes back to work.

What changed?

Nothing.

So I've started telling people:

Don't build an AI pilot that demonstrates AI. Build one that demonstrates competitive advantage.

• Did it identify a competitor vulnerability six months earlier?

• Did it recognize a product gap?

• Anticipate an acquisition?

• Detect a financial constraint?

• Help sales win a deal?

• Identify an inflection point while everyone else was still explaining yesterday?

That's competitive advantage.

And that's ultimately how GenAI should be measured.

From Information Collector to Intelligence Orchestrator

GenAI will consume a substantial amount of the information collection and basic analysis traditionally performed by CI practitioners.

Good.

Because collecting information was never the highest-value part of Competitive Intelligence. Judgment was.

The future practitioner will spend less time searching and more time interpreting.

Less time asking: “What happened?”

More time asking: “Why does it matter?”

Less time producing reports.  More time pressure-testing decisions.

We move from information collectors to intelligence orchestrators.

That's where AIAI, SAICI, and their CGEN framework become particularly interesting: putting methodology, governance, repeatability, and human expertise around GenAI so it becomes an intelligence capability—not another AI experiment.

But there's a catch:

You cannot orchestrate what you don't understand.

GenAI won't replace Competitive Intelligence.

But it may expose people who never really learned how to do it.

The future isn't human intelligence versus artificial intelligence.

It's human judgment multiplied by artificial intelligence.

Learn the trade.

Learn the tools.

Learn the methodology.

Then put them together.

Because when everyone has access to essentially the same GenAI, the technology itself isn't the competitive advantage.

Knowing what to do with it is.

Jay Nakagawa is recognized as a global thought leader and keynote speaker in Strategic Competitive Intelligence, Jay is passionate about helping organizations see around corners and make smarter, faster decisions in complex markets. His website is www.JayNakagawa.com.