Yes, AI will redefine marketing – but not the way you’d expect

Show me someone who says AI is game over for a profession, and I’ll show you someone who knows nothing about that industry and isn’t burdened by curiosity. The more some people know about technology, the less they seem to understand reality. People don’t buy legal contracts; they buy advice. They don’t buy solutions from McKinsey; they buy career insurance – or plausible deniability.

We see countless AI enthusiasts drifting near the industry, thinking advertising is just words on pictures or that marketing is just endless optimization of optimization. We’ve shifted to “vibe marketing” – make 500 ads, triple down on the ones people click, and think no further.

People act like being able to generate images for $0 instead of $1,000 is a game changer. But the image is only one tiny part of an execution being distributed with $500,000 of media behind it. Maybe the bit you’ve economized on matters more than you think.

Meanwhile, a LinkedIn post last week credited no fewer than 43 people for a single ad. This is an industry where spending seven figures on a photo shoot or agonizing over T-shirt colors for weeks is still a thing, just about. But those excesses, sensibly, are being recast as unsustainable.

Make no mistake: AI will have a profound impact on advertising, marketing, branding, design, CRM, media and every corner of business. But not in one leap. And not in the way most people expect.

Stage one: AI as a sparkly magic wand

This is where we are now. Tech people who don’t know the industry apply AI to specific, isolated tasks. In a rush to be first, vendors go narrow. In a rush to understand, clients want tools that require no integration and feel familiar.

So we get:

• AI to make nice images.

• AI to write nice copy.

• AI to optimize CTRs.

• AI chatbots (but now rebranded as “AI agents”).

• AI to cut mood films, make meeting notes, manage projects, polish animatics.

This phase brings hype, confusion, false promises, and a lot of reinventing the wheel with new labels. We’re not transforming processes – we’re slapping “AI” on stuff we used to just… do.

We see creative tools get democratized, yes – but mostly to raise the floor, not the ceiling. We see IP theft, hallucinations, and security disasters. We see businesses fire entire teams to look forward-thinking, only to quietly backtrack once the damage is done.

We get ads that technically meet the brief, but lack soul. A car commercial, made in an evening by people who don’t know ads are supposed to differentiate and seduce.

The result?

Beige.

Stage two: AI as a new transmission for old engines

A few years from now, the marketers and technologists start talking properly. Clients restructure. Teams get rebuilt. Real change takes shape.

This is where AI, and agentic AI in particular, becomes more than a bolt-on. It’s integrated into systems. We stop thinking about tasks and start thinking about transformation.

Now we get:

• CMO dashboards with clean, live data supporting better decisions.

• Agency operating systems orchestrating projects with more intelligence.

• Marketing production lines where ideation, creation, and optimization are streamlined and semi-automated – but still inspired.

This is where the human + AI loop delivers actual value. It’s no longer about “tools” – it’s about “how things get done.”

Stage three: AI as the blueprint for a new world

Eventually, maybe a decade from now, we will realize that AI hasn’t just changed the how but the what. Business models will shift, consumer behaviors will change, and the entire game will reset.

What was once hard or expensive becomes fast, easy, and cheap. AI-first businesses start competing with digital-first and traditional ones, both of which have had to adapt.

We’ll start rethinking marketing from first principles. What does a brand mean in this world? What can’t we measure, and why does it still matter? We’ll finally get honest about the limits of metrics – and the mysteries of persuasion.

New forms of communication will emerge, along with new roles, structures, and services. Marketing will become more imaginative and less operational, more about magic and less about implementation.

And if we want to steer this future, we need better questions – questions like:

• What is AI actually for in our organization?

• What level of decision-making should we apply it to?

• What risks are we not talking about?

• What timelines make sense for real change?

• Will AI work around us, or will we end up working around it?

• What parts of marketing still work – and which were always fiction?

Big questions. All worth returning to in future columns in The Drum.

Tom Goodwin is a globally recognized expert in marketing trends, AI, and digital transformation. He is the founder of All We Have Is Now, author of Digital Darwinism, and a keynote speaker at events all over the world. His work emphasizes “nowism,” a pragmatic approach to futurism that advocates using existing technology to drive meaningful change.

Connect with Tom on LinkedIn.

Read more opinion on The Drum.