I'm sitting in a meeting with the marketing director of a large FMCG company. He tells me their ad agency has "deployed AI." On the agency's website it says: "AI in our DNA." In the proposal: "services powered by artificial intelligence." In the presentation — a case study with generated graphics and copy.
I ask: "In which specific process does the agency use AI? Who owns that process? How do you measure the effect?"
The director looks at me and says: "I don't know. I think someone there uses ChatGPT to write copy."
And that's the crux of the problem.
"We Have AI" Is Not a Deployment — It's a Marketing Slogan
The advertising and marketing industry is going through a phase best described as AI-washing. Agencies are tripping over each other to communicate: "we're AI-first," "generative AI is part of our process," "we use AI in every project."
But when you ask for details, the picture changes. There are no modified processes. No sophisticated workflows. No deployment strategy that flows from agency leadership down to daily operations.
What's there instead? Someone on the creative team uses ChatGPT to generate first-draft copy. Someone in design experiments with Midjourney or DALL-E. Someone fed briefs to Claude to process them faster.
That's not an AI deployment. That's point-wise use of tools — exactly as an employee in any other industry might use a calculator. The tool is there, but the process hasn't changed. The workflow is the same. Work quality still depends 100% on manual human effort.
What "AI Deployment" in an Agency Actually Means
A genuine AI deployment in an ad agency would mean changes at three levels:
Strategy. Agency leadership defines where AI creates value — in the creative process, in analytics, in campaign optimization, in project management. Not "everywhere." In specific places, with measurable goals.
Tactics. Each defined area has a workflow where AI is embedded as a process element — not as an optional add-on. Roles are defined: who prompts, who verifies, who is responsible for output quality. There are quality standards. There is quality control.
Operations. Daily tasks change — not in theory, but in practice. Briefing looks different. Copywriting looks different. Reporting looks different. People work with AI, not alongside AI.
How many ad agencies operate this way? From my experience — very few. Most stopped at "someone here uses ChatGPT" and turned that into a sales pitch.
Why This Should Matter to Agency Clients
If you're a marketing director buying services from an agency that markets "AI," you should ask a few questions:
- In which specific process do you use AI?
- Who owns that process?
- How did your workflow change after deploying AI?
- How do you measure the effect — speed, quality, cost?
- Is AI part of the standard process, or an optional experiment?
If the answers are things like "it depends on the project" or "our team decides on their own when to use AI" — that's not a deployment. That's a gadget.
And you shouldn't be paying an "AI" premium in the price of services for a gadget.
A Prompter Is Not a Strategist
There's a difference between an agency that has a "prompter" — someone who types text into ChatGPT — and an agency that has a strategy for using AI in the creative process.
A prompter does what a copywriter did, just faster. A strategist changes how the agency creates value.
A prompter generates a first draft. A strategist asks: why are we writing this text manually at all? Can the entire briefing-copy-review-approval pipeline be redesigned? Where does AI shorten the path? Where is a human irreplaceable? How do we measure whether the new process is better than the old one?
Companies that truly want to leverage AI in marketing should look for partners who understand this difference. And they should be prepared to hear that their current agency — despite beautiful declarations — may not understand it.
From Slogan to Process
I'm not saying ad agencies deliberately deceive clients. Most genuinely believe they "use AI." The problem is that the barrier to entry is so low — a ChatGPT account is all it takes — that it's easy to confuse a tool with a transformation.
Real change requires something the creative industry has historically disliked: processes, standards, measuring outcomes. AI doesn't tolerate creative chaos — it needs structure. And structure has to be deliberately designed.
If your company uses ad agency services and you want to understand whether there's real change behind the "AI" label — a process audit will show where there's value and where there's just marketing.
Dear Reader, if you want to discuss how to assess the real level of AI deployment at your partners and suppliers — I invite you to get in touch. Leszek Giza.