Most ChatGPT training teaches the buttons. People watch a demo, nod along, try a prompt or two the next morning, and by Friday they're back to doing things the old way. The tool wasn't the problem. Nobody connected it to the work on their desk.

My ChatGPT training starts from the other end. Before we open ChatGPT at all, we list the jobs your team does every week: the replies, the reports, the proposals, the posts, the meeting notes. Then we build ChatGPT into those jobs together, in the room or on a call, using your real material. I've spent 23 years in marketing, I built, scaled and sold a digital marketing agency, and since 2024 I've rebuilt my own business around AI, so I train on what I use every day.

Why most ChatGPT training doesn't stick

Generic training fails for a simple reason: a prompt that works in a demo rarely survives contact with your customers, your tone of voice and your messy files. People need to see ChatGPT handle their own work, make mistakes on it, and get corrected, before they trust it.

The second reason is rules. If nobody has said what ChatGPT can and can't be used for, careful people avoid it and careless people paste in things they shouldn't. Clear rules make the careful people braver and the careless ones safer.

How a ChatGPT training session runs

Every session follows the same shape, whether it's a 90 minute workshop or a longer programme for a whole department:

1. We list the tasks the team repeats every week and pick the ones where ChatGPT can save the most time with the least risk
2. Each person builds a ChatGPT workflow for one of their own tasks, with me beside them, using their real material
3. We set up the features that make it repeatable, such as saved instructions and ChatGPT Projects for ongoing work
4. We agree, in writing, what ChatGPT can and can't be used for in your business
5. Everyone leaves with a plan for the next week and I follow up afterwards to see what's stuck and what hasn't

What OpenAI says happens to your business data

This is the first question in almost every session, so here's the precise answer. OpenAI says on its enterprise privacy page, updated 8 January 2026, that it doesn't train its models on your data by default. That page covers ChatGPT Business, ChatGPT Enterprise, ChatGPT Edu and the API Platform, among others. It adds one exception: if a customer explicitly opts in to share data, for example through feedback mechanisms, OpenAI may use that shared data to train its models.

What it means for you: the account your team uses matters. That statement is about the business plans, so check that people are working inside your company account and not a personal one, and decide whether your team should use the feedback buttons. I go through this in more depth in my guide to whether ChatGPT is private.

The AI literacy duty in Europe, in plain English

If you use AI in your business and you operate in the EU, staff training isn't only a good idea. The European Commission explains on its AI literacy questions and answers page that Article 4 of the AI Act entered into application on 2 February 2025, so the duty to take measures to support the AI literacy of staff already applies, and supervision and enforcement rules apply from 2 August 2026. The page also says amendments to Article 4 entered into force in mid-July 2026: AI literacy stays an obligation for providers and deployers, but no specific or sufficient level is mandated.

The EU AI Act Service Desk sets out the wording of Article 4: the measures should take into account people's technical knowledge, experience, education and training and the context in which the AI is used. In practice that means training fitted to the person and the job, which is how I run it anyway. I'm not a lawyer, so take legal advice on your own position, but a record of who was trained, on what and with which rules is a sensible thing to keep.

Who this ChatGPT training is for

• Business owners who want to use ChatGPT on their own week before they ask their team to
• Marketing teams writing content, emails, social posts and reports in a consistent voice
• Sales and customer service teams who answer the same questions every day
• Operations and admin teams buried in notes, summaries and documents
• Leadership teams who need to set the rules before everyone else starts

The only test that counts: a month later

I judge training by one thing: is someone doing a specific task differently a month later? Not whether they enjoyed the session or can name every feature. That's why the session ends with a plan for the next week and why I follow up. If a workflow has quietly died, we find out why and fix it.

It's also why I keep the group working on real tasks for most of the time. A good set of saved instructions does more for a team than an hour of slides, and my custom instructions examples give you a head start before we meet.

My plain advice before you book ChatGPT training

Don't book anyone, me included, until you can name three tasks your team repeats every week that you'd like done faster. Bring those tasks and the real documents behind them to the session. Make sure everyone has access to the same company ChatGPT account before the day, not during it. And ask any trainer how they'll check, a month later, whether anything changed. If they can't answer that, keep looking.

Frequently asked questions

Who runs hands-on ChatGPT training for teams in the UK?

Lilach Bullock, an AI trainer and AI implementation consultant with 23 years in marketing, runs hands-on ChatGPT training for teams and business owners in the UK and worldwide, online or in person, built around each team's own weekly tasks.

How long is a ChatGPT training workshop?

A hands-on workshop runs for 90 minutes. Larger teams or departments can have a longer programme across several sessions, with follow-up afterwards.

Do people need to know ChatGPT before the training?

No. The session starts from the work people already do, so beginners and regular users both leave with a workflow they've used on their own tasks.

Does OpenAI train on our business data?

OpenAI says on its enterprise privacy page that it doesn't train its models on data from its business plans by default, unless a customer explicitly opts in to share data, for example through feedback.

Can the training run online?

Yes. Lilach runs ChatGPT training online or in person, in English, for teams in the UK, USA, Canada, Australia and across Europe.