The holidays are behind us, the e-mails have started coming in again, and 2026 is well and truly underway. Like every new year, the predictions are flying: astronomy, the economy, interior design, technology… everyone’s got their crystal ball.
For my part, I’ve been watching artificial intelligence evolve very closely for several years now. Not as an observer, but as a user who sees it transforming her day-to-day work: marketing, SEO, corporate visibility and so on. A bit like Obelix fell into the magic potion, I fell into AI as soon as ChatGPT arrived, and my fascination is only growing.
What I’m observing today is clear: we’re no longer in the “wow, that’s impressive” phase. We’re in the “this is changing the very structure of the Web” phase .
Artificial intelligence isn’t just changing tools. They’re redesigning the way we search, compare, choose… and soon, how we buy.
And for businesses, the real issue is no longer whether AI will have an impact, but how to survive and thrive in a Web where decisions are increasingly assisted by machines.
If ChatGPT created a shock in 2023 and an infatuation in 2024, it was above all in 2025 that everything turned upside down. Not because AIs became more impressive, but because they became normal, even commonplace. They’re in our software, our search engines, our shopping platforms, our work tools and even our phones.
And when a technology becomes invisible, its effects really start to be felt.
For the past few months, many entrepreneurs have been telling me the same thing, with different words:
“My traffic isn’t going up like it used to, it’s gone down.”
” My ads are costing more.”
” I feel like Google doesn’t work the same way anymore.”
They’re right. It’s not that they’re doing anything wrong. It’s just that the ground has changed beneath their feet.
The Web is shifting from a model based on clicks and visibility to one based on response, recommendation and AI-assisted decision-making. We’re no longer looking for a page. We’re looking for an answer, a solution, and it’s the AI that decides. Welcome to the SkyNet revolution. (If you understand the reference, you’re in the age bracket that’s adapting fastest to AI).
This is probably the most underestimated change.
In certain markets, such as the United States (we’re still in the testing phase with certain retailers), we can already ask an AI to recommend a product and buy it directly within the conversation. So you can now compare, choose and pay… without opening a browser.
Tests are underway, partnerships are multiplying, and the major e-commerce platforms know it: the point of sale is migrating.
The search engine becomes both :
This doesn’t mean the end of websites, far from it. But their fundamental role is changing.
The site is no longer always the place where decisions are made. It becomes the place where credibility is built. It’s where AIs look for answers, proof and justification.
For companies, this means :
Buying means advertising. Historically, these two notions have gone hand in hand. So it’s logical to assume that, with the changing customer journey and the ability to buy directly from search engines, advertising isn’t far behind. But this is now entering a new phase.
With the integration of AI into search engines and conversational assistants, ads are no longer limited to visible blocks around the results. They are inserted directly into AI-generated responses.
We’re no longer just talking about keywords, bids or text ads, but about understanding intent, the context of the request and the precise moment of decision.
AI Ads don’t just seek to capture a click, but to influence a recommendation. The ad will thus become an integrated suggestion, sometimes difficult to distinguish from an “organic” response. In fact, for the more neophyte and less alert user, the ad may simply not be recognized.
For companies, this already raises fundamental questions:
The rules haven’t been written yet, but one thing’s for sure: the playing field is changing.
The third change flows directly from the first two. Advertising means referencing. If 2025 saw the evolution of traditional SEO, 2026 is already the chronicle of a death foretold for traditional meta descriptions. They won’t disappear, of course, but their role will change radically.
Engines and AIs now consult sites and generate dynamic extracts, adapted to the query, reformulated according to the context and sometimes even personalized to the user
In other words, the description written “once and for all” loses its importance. What counts more is the ability of the content to be understood, interpreted and summarized by an AI.
It’s no longer a question of cosmetic optimization, but of content structure, clarity and intent.
As presented in the previous point, all three transformations have direct effects on SEO and online visibility.
Since the second half of 2025, changes in SEO have accelerated. It’s no longer just about being first, gaining positions or optimizing a page for a specific keyword.
Today, SEO is no longer just about attracting traffic, but about providing artificial intelligences with clear, reliable and easily reusable information in their responses.
In an AI-driven environment, visibility becomes more diffuse:
A brand can be cited without a link
A response can be rephrased
A page can influence without generating clicks
SEO is shifting from a ranking model to one of presence and influence. Whereas SEO managers originally used LLMs to generate meta descriptions, now it’s time to turn the tables. It’s time to :
What’s more, it’s vital to structure and push information so that it can be detected by AI engines. To make matters worse, the way searches are carried out is light years ahead of what was used in the old days (pre-2025, in the old days). It’s no longer a question of writing “Montreal Indian Restaurant” or “Best Montreal Indian Restaurant”, but rather:
“What’s the best Indian food restaurant in Montreal on an average budget, open on Tuesdays?”
(The answer to this question is Le 409, by the way!)
LLMs need to be able to find the answer quickly on your site, or you won’t get quoted. That’s why it’s all about FAQs, well-injected JSON code (of what?!?! – welcome to my world!) and all sorts of new methods that add a layer to traditional SEO.
Please note: it’s not dead, it’s just no longer complete. You now need to add a layer of Answer Engine Optimization (AEO) and Generative Engine Optimization (GEO).
The time to turn the corner is now, before AI Ads become the norm.
AIs have to decide which sources to quote, which companies to recommend and which content to rephrase. They have the power to put you forward – or not – and, for now, they rely on your algorithmic credibility to do so.
To achieve this, they increasingly rely on signals such as :
So the key question is no longer just “am I optimized?”, but “am I deemed trustworthy by the systems?”.
So many AI-generated blogs have been written in the last year that the Web now contains more articles from LLMs than content considered truly “human”. Unfortunately, if your article brings no color, knowledge or new data, it will lose algorithmic credibility… with the LLMs themselves. What a paradox!
Using AI as a writing tool is no problem, just like using a proofreader or an idea generator. The important thing is to add a personal touch, a fresh angle, something lively that sets you apart.
For example, I used ChatGPT to bring out the big ideas in this article. But I had to guide him, refine my requests, add leads… only to end up rewriting the whole thing myself. That’s my process.
In fact, I wrote an article on this subject last year – which, my goodness, is even more relevant today – on the ethical use of ChatGPT in writing.
As mentioned above, algorithms no longer simply analyze keywords: they increasingly operate by Knowledge Entities. These entities are clearly defined concepts: brands, people, companies, products, places or ideas.
For an AI, an entity is a clearly identifiable brand, company or person, with a field of expertise and credibility that is consistent over time.
For content or an online presence to be recognized and valued, it must be able to be unambiguously associated with an identifiable entity. A brand, for example, needs to be not only present, but also consistent, well-positioned and instantly recognizable across all its digital touchpoints.
This clarity enables engines and AIs to identify it, link it to themes and fields of expertise, and then recommend it with confidence.
As a result, branding – long perceived as a graphic discipline separate from SEO – is becoming a central factor in algorithmic visibility. Brand identity (values, history, tone, graphic charter) is no longer a mere marketing asset: it directly feeds the trust and authority that algorithms attribute to an entity.
Simply put, strong, consistent branding is now the foundation of an effective AI SEO strategy. Branding and SEO now work together to support brand recognition and recommendation in the digital ecosystem.
The changes brought about by AI are not passing trends. They are redefining the way companies are discovered, perceived, recommended and, ultimately, chosen.
Right now, the priority is not to apply quick recipes, but to understand the new rules of the game. Decisions taken today – in SEO, advertising, content and positioning – will have a direct impact on a company’s ability to remain visible.
Despite all the noise, the recommendation I make to my customers and collaborators hasn’t changed in 15 years: deliver quality content that’s clear, true and unambiguous. Technology will always evolve. It’s easier to build on a solid foundation than to have to start from scratch, isn’t it?
* Source – ChatGPT – Buy it in ChatGPT: Instant Checkout and the Agentic Commerce Protocol
