Right now, someone in your company is pasting client data into ChatGPT
Most CEOs of mid-sized companies do not know it is happening. Most will not, until it becomes a problem with a name, a date, and a legal notice attached.
Field notes
Everyone says 2026 is the year AI changes everything. You are still in committee.
Morgan Stanley told institutional money to brace for something. GPT-5.4 is already at expert level on economically valuable tasks. The curve gets steeper from here.
A five-person company is going to outcompete you in 2026. Here is how.
Sam Altman's near-term vision: companies of one to five people outcompeting large incumbents. This is not science fiction. It is a description of something already happening.
There are four stages. Most companies are stuck in stage one.
Most companies are in stage one and calling it a strategy. Here is what the four stages actually look like, and what separates the ones moving through them.
Your AI project is not failing because of the technology.
If there is clarity and capable teams, it accelerates. If there is organizational disorder, it exposes it. Most companies assume they are in the first category.
You adopted AI. That is not the same as implementing it.
More than half of companies use AI somewhere. Almost none have significant financial impact from it. The gap between those two things is not a technology problem.
Your best people are funding AI out of their own pockets. Your AI budget is still under review.
The ROI was not demonstrated through a corporate pilot. It was demonstrated through thousands of individuals making the same quiet decision: this is worth twenty dollars of my own money.
Glean alternatives: open-source enterprise AI search without the €50k/year lock-in
Glean is the gold standard for a reason. But the price tag excludes most mid-market companies. Onyx and the managed open-source model change that equation — if you know what you are trading.
Are you an AI company, or a company that has an AI slide in the deck?
Are you an AI company, or are you a software company with AI features? The distinction matters more than it might seem right now.
You are trying to buy AI the same way you bought software
The CEO is not being slow or cautious. The CEO is applying a rational procurement process to something the procurement process was not built to handle.
What does an AI readiness assessment actually cost in 2026?
Bpifrance funds them at €10k for eight days. Big Four price them at five times that. What you actually get depends on who does the work and what happens after the report.