OpenAI passed 1 billion active users per week this month. And since this summer, there has been an advertisement among the answers.
Those are two facts that, as an employer, you cannot view in isolation from each other.
Candidates no longer put their career questions to Google. They put them to ChatGPT. “Which employer suits someone like me.” “Where can I really progress.” “Which company has a proper culture.” The answer they get back determines who is on the shortlist, before you have ever seen that candidate.
And that answer is from now on partly for sale.
The ad rollout: Europe's turn
OpenAI started with ads in the United States, Canada, Australia and New Zealand. In May, the United Kingdom and Japan were added. This month, 31 European markets will follow, including Germany, France and Spain. Advertising revenue grew by more than 25 percent in a few weeks, heading towards an annual revenue of nearly 1 billion dollars. One in five users shows commercial intent. It is a business model that is taking root.
The rollout is marching across Europe. Whether the Netherlands is up this month or next month matters little. The direction is clear.
No more click-throughs to your careers site
Meanwhile, the world in which a candidate clicks through to your careers site is disappearing. Research by SparkToro, based on clickstream data from Similarweb, showed that 68 per cent of Google searches in the early months of 2026 ended without a single click. In 2024, that figure was 60 per cent. And if there is an AI summary at the top, the likelihood of a click-through drops by almost 60 per cent.
Translate that to your recruitment process. The candidate reads the answer. The candidate forms an opinion. The candidate does not click.
Your careers page, your EVP video, your Glassdoor reviews: they still exist. Only, candidates are visiting them less and less frequently. The verdict is reached a step earlier, in a conversation you are not a part of.
A Google list gave ten blue links and left the choice to the candidate. An AI answer already makes that choice in part. It names three employers, with a reason included, and leaves the rest out. There is no second page. Whoever is not in that first set does not exist in that conversation.
For recruitment teams, this feels unfair, and it is. You have spent years building an employer brand, campaigns, and a narrative that rings true. And now a model you have never spoken to decides whether that story reaches the candidate. The good news is that that model has no opinion. It reads what is there. And what is there, you can steer.
For employers, the playing field is now shifting on two fronts simultaneously.
The first front is organic. Whether you are mentioned when an AI advises someone about employers in your sector, or whether you fall silent. That depends on whether your information is consistent and citable. On whether language models know your brand as an entity with a story, or as an empty shell with a vacancy feed.
The second front has been paid for. Soon you'll buy a place in that answer, just like you buy Google Ads now. Fine, until everyone does it and the price goes through the roof.
Look at what happened to Google Ads. The first advertisers paid pennies. Ten years later, a recruiter pays tens of euros for a single click on a scarce vacancy. An advertising market rewards those who are early and punishes those who arrive late. In AI, that clock is now starting to tick.
Here is the calculation that most executive boards haven't made yet. Whoever is organically invisible will soon pay top dollar just to be mentioned at all. Whoever builds their visibility in language models now will soon have a foundation that no advertising budget can replace.
What is optional today will be an advantage in six months. What is an advantage in six months will be the baseline in a year. And once it is the baseline, you cannot buy your way back with a campaign.
What does that mean in practical terms, this week.
Find out what AI is saying about you today. Ask the questions your candidates are asking ChatGPT, Gemini and Perplexity, and read the answer the way a job applicant reads it. Do it systematically, every single month. You will be more shocked by the silence than by a wrong answer.
Make sure your information is consistent everywhere. A language model that reads something different on your site than it does on LinkedIn, and something different again on Glassdoor, will play it safe and leave you out. What is unclear gets simplified. What is not consistent anywhere does not exist.
Treat your employer brand as a source that AI can cite. Ads in AI are bought per click. Authority in AI is built up. The first stops as soon as your budget stops. The second remains.
The ad is coming, whether you are ready or not. The question is whether your brand is strong enough to appear in the answer without that ad.
My name is Bryan Peereboom, founder of AI Rebels. We make employers visible in language models.