AI in marketing: What your CFO is excited about... and should make your CMO uncomfortable

L'Oréal, Walmart and the wave of “savings with AI” tell a different story if you read them from a strategy perspective, not just from efficiency.

In 2025, the conversation about AI in marketing was filled with promise: faster, lower costs, more “intelligent automation”.

But if you look calmly at some of the recent news, another story emerges: it's not that AI replaces agencies and teams, it's that Make it clear who has a strategy and who only had a content calendar.

Three examples:

• L'Oréal testing “AI filmmakers” for campaigns.
• Walmart, Amazon and Sephora using AI to guide buying.
• The pressure to “save with AI” collected by media such as Wall Street Journal and controversies over poorly managed automated campaigns.

Add to this what the global reports on digital use say: mobile and social continue to rule, while sensitivity to how brands use data and AI grows.

With this context, the question isn't “am I going to use AI?” , if not: What happens to my marketing when AI comes into play?

1. L'Oréal and the “AI filmmakers”: more pieces don't always mean better marketing

According to The Times, L'Oréal is working with teams of “AI filmmakers” capable of generating video pieces and campaign creatives without going through a traditional set.

This has two possible readings:

• The simplistic reading: “if L'Oréal uses AI to create content, so do I; that way I produce more pieces at a lower cost”.
• The useful reading: “if I'm not very clear about who I am as a brand, what visual language is mine and which isn't, AI will only produce variations of something that is already diffuse”.

Technology multiplies production. Identity and judgment remain human responsibility.

2. Walmart, Amazon, Sephora: the decision moves to the conversation

Business Insider describes how Walmart, Amazon and Sephora integrate AI into the shopping experience: assistants who recommend products, assemble baskets and simplify the checkout.

That means that:

• The user doesn't always go to “tour your site”; they ask a system what to buy.
• The difference between your brand and the one next door is increasingly decided in How do they describe your products and what information those systems have about you.
• Your UX is no longer just that of the site, but that of all the interfaces where the decision is made (including the attendees).

If your digital roadmap is still focused solely on “renewing the homepage” and on traditional campaigns, there's a part of the funnel that you're not even seeing.

3. Side B: “savings with AI” and campaigns that go wrong

In parallel, different notes collected by means such as Wall Street Journal And Business Insider they talk about two trends:

• managers who pressure marketing to justify cuts thanks to AI (“if we automate, why do we need the same equipment?”) , and
• campaigns that rely excessively on automatic generation that end in criticism, confusion or reputational damage.

It's no accident: when AI is used only as a tool to produce faster and cheaper, without a strategy behind it, it amplifies noise and risk.

4. AI as an amplifier of strategy (or lack of it)

Connecting this news to the agenda of a medium/large company leads to a simple point:

• With a solid strategy, AI helps to better investigate, test hypotheses, automate routine tasks and manufacture variants that were previously unfeasible.
• Without strategy, AI only generates more things, faster, with the same level of confusion that already existed.

That's why the conversation shouldn't be “AI yes or no”, but rather:

• What decisions do we want to keep in the hands of people (positioning, priorities, narrative, experience design)?
• At what points in the process do we want AI to support us (research, analysis, production, operations)?
• Who within the team has the competencies and the mandate to connect both layers?

5. A reasonable stance against AI

From undertk, the position that can be communicated (without selling, only as a criterion) is something like this:

• Use AI A lot and well for what it does best: research, order, detect patterns, automate repetitive things, simulate scenarios.
• Don't delegate to AI what defines brand value: the underlying strategy, the narrative, the decisions about what not to do, and the design of the business and user experience.
• Accept that AI changes the context (more speed, more complexity), but not the responsibility to decide where the company moves.

It's a message that a serious CMO or Founder can respect: it's not “the machine replaces you”, it's “the machine forces you to take your work more seriously”.

6. Questions to review before buying another tool

If you want to use this news to act on something concrete, more useful than talking about “transformation” in the abstract, you can start with these questions:

1. About creativity
2. About digital product and UX
3. On Data, Metrics, and Trust

If the answers are weak, the problem isn't a lack of AI. It's a lack of strategy connected to UX, content and operation.

The news about L'Oréal, Walmart or the cuts “thanks to AI” are not instructions for dismantling teams and agencies. They are reminders that:

• AI is already part of the context.
• The difference is made by who takes the strategy seriously, not who installs the most tools.

And that's where it makes sense for a company to wonder if the way it's marketing today... is really up to what the context is asking of it.

‍