ChatGPT training for companies: workshop, half day or full day, online or on-site
Three formats, your own tasks on the screen.
Your team solves real work tasks in ChatGPT with us in the room, and two hours is enough for everyone to finish at least one.
What is ChatGPT training for companies?
The account is usually there already, and the open question is whether anyone uses it for more than rewriting emails. ChatGPT training for companies is a guided session where the team solves its own work tasks in the tool with us sitting next to them, instead of watching slides about what AI might do one day.
We use the same picture as in our Claude training: ChatGPT is a capable intern on day one, who follows instructions to the letter, guesses when something is unclear and needs to see what a good result looks like. Once that picture lands, people suddenly know how to describe a task and how to check the answer.
Since 2026 the training also covers ChatGPT Work, the agent that takes a goal and hands back a finished spreadsheet or document. A controller in Rotterdam can ask it to read twelve supplier invoices and build a reconciliation tab, although someone still reads the result line by line before it goes to the auditor, and we practise that part too.
Three formats for your team
The same core runs through all three, so choose by how much time you have and how far the team has already come.
ChatGPT workshop, 2 hours
For teams that already have licences but where the admin panel shows half the seats never log in. Everyone solves one real task and leaves knowing where the line for customer data runs.
- First: The intern metaphor and a first task from each person’s own week.
- Then: The task brief: goal, input, format and an example of the finished result.
- Last: The line: what may be typed in on your plan, and why national ID numbers and customer names stay out until someone has read the agreement.
ChatGPT workshop, half day
Everything in the two-hour format plus group exercises per role. It fits a team that already uses ChatGPT every day, only in ten different ways.
- Basics: The intern metaphor, the task brief and the data line, as in the two-hour format.
- ChatGPT Work: The team hands a whole task to the agent and sees where the usage limit of a seat sits.
- Per role: Group exercises built on tasks from your own work.
ChatGPT workshop, full day
The morning is a shared foundation. The afternoon splits so marketing, finance, HR and leadership each get their own track with their own tasks.
- Morning: The same foundation as the half day, with ChatGPT Work on your own files.
- Afternoon: Separate tracks for marketing, finance, HR and leadership.
Prices are per user with at least 8 participants, exclude VAT and are the same whether the team runs ChatGPT or Claude. The licence comes on top: OpenAI lists a Standard seat on ChatGPT Business at US$20 per user per month on annual billing.
Does your team lack shared ChatGPT seats?
OpenAI sells self-serve ChatGPT Business by card only, and an invoice needs a contracted plan such as Enterprise. That is why accounts so often sit on private cards, where nobody can see who actually uses them.
We buy the seats for you, set up the workspace and invoice in euros once a month: €32 licence and €14 operations per user and month, plus €490 once for the setup. The training and the agreement then come from the same place.
How do you choose the right format?
Choose by the admin panel, not by the budget. If fewer than half of the seats are used each week, two hours is enough, while a team that uses it daily but all over the place gets more out of the half day, and a whole department with mixed roles gets the most out of the full day.
The format matters less than the week after. What separates teams that keep going from teams that stop is a fixed one-hour sharing slot every week, so book the follow-up fourteen days later at the same moment you book the training, or it never happens.
What can the team do after two hours?
Every participant has solved at least one task from their own work, for example turning forty minutes of meeting notes into three decisions and five actions, each with an owner and a date.
We measure it in the room by counting how many runs needed correcting before the result could be sent on. ChatGPT writes a draft in ten seconds, although knowing whether it is right still takes a human, and that check is the part we spend the most time on.
Do staff have to take AI training?
Article 4 of the EU AI Act has applied since 2 February 2025 and asks companies that deploy AI systems to take measures so their staff understand what the tool does and where it falls short. The Digital Omnibus on AI, in force since mid-July 2026, keeps that obligation but no longer sets a specific level, so documented training with an attendance list and the content covered is still the simplest proof, and we hand that documentation over after every session.
The data question belongs in the training as well. ChatGPT Business lets the workspace owner choose Europe as the storage region, although OpenAI is rolling that out gradually and keeps a copy of prompts in the United States for a limited time, so we spend ten minutes of every workshop on which plan and region you actually have and what that means for what may be typed in.
ChatGPT or Claude in the training?
The teaching stays the same: the intern metaphor, the task brief and the line for what may be typed in. What changes is the tool on the screen and where its limits sit, and although we build most things on Claude from Anthropic, we run the same workshop in ChatGPT for teams that already have the OpenAI agreement.
If you are still weighing the two, decide before the training rather than during it, because otherwise the team learns two tools halfway and neither sticks.
Do this before the first session
Open the admin panel in ChatGPT Business and write down how many of the seats were used in the last thirty days. That single number is your starting point, and it is the one you compare against thirty days after the training.