Should I remove the word AI from my website?

Not necessarily. Remove the phrases that have no result for the client and no proof behind them. If AI makes the work faster or cheaper, say by how much, in hours or money, and mention the technology further down the page.

Claims about using AI that a company cannot back up. In the US it is already fined: in March 2024 the SEC charged two investment advisers a combined 400,000 dollars.

How do I check my landing page quickly?

Search the page for "AI", "artificial intelligence", "neural", "machine learning", and for each phrase answer three questions: what does the client get, who is responsible if the machine is wrong, and what proves it. Rewrite or remove phrases without answers.

The word "AI" on your landing page: three questions I ask every such phrase

On October 16, Journal du Net published my ninth column, about how the word “AI” stopped working in B2B marketing. For French readers I went through the numbers and the court cases. Here I will briefly say why, and show in more detail how to check your own site.

Back in 2024, researchers at Washington State University showed people identical product descriptions that differed only in whether artificial intelligence was mentioned, and where it was, people were less willing to buy. In April 2026, Emplifi asked more than 1,600 people in the US and the UK which content formats they trust: 31% trust content made by AI, 63% trust customer reviews. In B2B, according to Forrester, when a purchase includes generative AI, the group that makes the decision roughly doubles, and 28% of buyers said inaccurate information from AI made them less confident in their choice.

Empty “AI” claims are also starting to cost money. In March 2024 the SEC fined two investment advisers a combined 400,000 dollars for AI claims they could not back up. In May 2026 Apple agreed to pay 250 million dollars to settle a class action from iPhone buyers who were sold phones on Siri AI features that did not exist at the time of purchase (the company does not admit wrongdoing).

Now the check itself. I open the site, search the page for “AI”, “artificial”, “neural” and run every phrase I find through three questions.

What does the client get? If the honest answer is “nothing, we just use AI”, the phrase is empty. “AI-powered sales analytics” becomes “a report on every lead, every morning”. And if AI really makes the work faster or cheaper, I write by how much, in hours or money. The technology can be mentioned further down, for those who want to know how it works.

Who answers when the machine is wrong? In B2B this is the main fear, and what removes it is not a description of the algorithm but a person: the name, photo and phone number of whoever checks the result. The more a mistake costs the buyer, the higher on the page this should be.

What proves it? “AI shortens delivery times” - by how much, and on which project? Nothing to show - the phrase goes. A couple of years ago nobody would have noticed it. Now clients notice it, and so do regulators.

Clients care about what the work costs, when the result arrives and who answers for it, not what it was made with. AI is the same story as the cloud, which fifteen years ago sites listed as an advantage and today nobody mentions.

The full column, with sources, is on Journal du Net, in French.

Have you removed “AI” from your own copy? I’m curious what happened to your leads.

Source of the story: my Journal du Net column, “Le mot IA ne fait plus vendre. Comment parler à vos clients B2B quand la hype retombe”, October 16, 2026.

Sources of the figures:

Washington State University, press release, July 30, 2024 (Journal of Hospitality Marketing & Management) - mentioning AI lowers purchase intent.
Emplifi, content trust study, US and UK, 1,650 respondents, April 27, 2026.
Forrester, State of Business Buying 2026, via Digital Commerce 360, January 22, 2026.
SEC, press release 2024-36, March 18, 2024.
MacRumors, May 5, 2026, and AppleInsider, July 24, 2026 - the Apple settlement.

Frequently asked questions