.svg)
Over the past few months, you've seen it appear everywhere: AI-driven agency. AI-first approach. We're AI powered.
Some agencies even put it pontifically on their homepage, as if it were a quality label. Like you can't be taken seriously today without the word AI in capital letters.
We get it. AI is important, the impact on our lives is huge :)
The technology is developing rapidly and will change a lot in the coming years. But that is precisely why it is remarkable how quickly some organizations act as if the outcome is already certain. As if “AI-first” is the logical destination. But it isn't. In fact, the term already shows signs of aging.
The illusion of acceleration
The problem with an “AI-First” principle is that it creates the illusion of a “right” solution to every problem you have. In this way, real critical questions about value and return are often parried or postponed. As soon as an agency or organization defines itself on the basis of a technology, the conversation shifts almost automatically to implementation and tooling, instead of to the key question: what does this actually give me, and when?
The questions that stakeholders ask are rarely abstract. They are practical, urgent and contextual. What if I don't want or need AI at all? As a manufacturer of toothbrushes, a manufacturer of clothing or a distributor of food, how is AI going to help me grow, better serve customers and earn money? Can I invest in something that only pays off in six months or later?
These are not technological questions, but business questions. And that's exactly where the “AI-first” thinking comes in. It suggests an obvious value, while in practice, that value always depends on context, timing and scale.
In addition, many AI projects in reality do not accelerate, but slow down. Research by McKinsey (2024) and MIT Sloan Management Review (2025) shows that large-scale AI deployments take an average of 12 to 18 months before they produce demonstrable results. During that period, time, capital and attention are committed, while operational improvements often come to a standstill or are postponed. Competitors who do not fully focus on AI, but continue to optimize within existing structures, regularly appear to move faster in that phase.
The promise of acceleration thus masks an uncomfortable reality: AI can create value, but requires delay, restructuring and investment first. Ignoring that is not selling a vision, but a simplified representation of reality.
Why this breakthrough is less new than it looks
The current wave of AI fits into a pattern we've seen before. Mobile interfaces, social media, cloud computing, and big data were all presented as fundamental breaks with the past. During those periods, many agencies positioned themselves as “mobile-first”, “social-first” or “data-driven”, often without this being deeply translated into strategy, organization or business model. Afterwards, those labels turned out to be mainly marketing, not structural choices.
AI follows a similar path. The visibility is new, the attention is explosive, but the underlying development has been ongoing for years. Most of the applications that are effective today are not about full autonomy, but about support: AI that analyses, suggests, signals and automates within existing processes. Man and machine work together, with human judgment remaining leading.
This makes the current positioning as “AI-first” all the more striking. It suggests an end state that is still far away in practice and misses that technology rarely develops linearly or predictably. At the same time, it raises questions about timing and credibility. Organizations that are now suddenly profiling themselves with terms such as AI-first or AI-driven are visibly responding to a wave of attention, while AI as a technology has been part of research, tooling and operational systems for years.
If AI was really so fundamental to the way they work, that shift should have occurred earlier and more consistently in their services, propositions, and choices. Not as a label, but as a silent, structural development. The late and emphatic embrace of this terminology therefore feels less like a long-thought-out vision and more like positioning: an attempt to stay relevant in a debate that has suddenly become public and commercially attractive.
The conversation that is structurally avoided
What is conspicuously missing in many AI discussions is an open conversation about choices, risks and timing. About what an organization is willing to invest, what that means for people and processes, and what assumptions lie beneath that investment. Ethical issues and transparency towards customers and employees also often remain understudied, while social tension around AI is increasing there.
Once an organization publicly calls itself “AI-first”, that conversation becomes more difficult. The identity is fixed, the promise has been made. Reticence or doubt then feels like weakness, while in reality it often indicates careful consideration. You get caught up in a technological narrative that doesn't necessarily match your business model, your customers, or the reality of your market.
The reality of adoption
AI will gradually settle into organizations in the coming years, following the pattern described by Everett Rogers in his Diffusion of Innovation theory: innovators, early adopters, early majority, and ultimately the late majority. Nobody gets fully AI-driven overnight, which is exactly why it's risky to position yourself like that right now.
The organizations that are relevant ten years from now will not be the ones with the loudest AI story, but those who use technology where it demonstrably adds value. Where AI supports employees, strengthens decision-making, simplifies processes and improves customer experiences. Not as a label, but as part of a broader, realistic development.
Conclusion
“AI-first” sounds progressive, but says little about what an organization actually achieves. It can provide a false sense of security, lock in resources without immediate impact, and delay the conversation about true value.
AI is important. AI will have a huge impact in the coming years. But it remains a technology, not an end goal. The real gain is not how quickly you position yourself, but in how carefully, context-oriented and realistic you make choices. That's where value is created that stays, even when the slogans and expletives have disappeared.
