In brief
  • Most AI training budgets buy one-day events that create enthusiasm and no lasting change. What you are actually buying is adoption: a change you sustain, not an event you attend.
  • AI adoption succeeds or fails on leadership. If leaders are not invested, present and visibly using the tools themselves, people will quietly resist.
  • Choosing the partner is not a procurement task to delegate. The most senior people who will own the outcome belong in the room from the first conversation.
  • A strong inquiry names your teams, tools, outcome, constraints, budget range, success measures and decision-makers. There is a copy-ready checklist below.
  • Test every proposal against one question: will this partner still be useful to us in a year?
In this guide

If you lead learning and development, this is the season when the AI training inquiries go out. Budgets for next year are taking shape, the board wants a credible answer on AI capability, and your inbox is filling with providers offering to transform your workforce in a day.

I sit on the other side of those inquiries, and I will be direct with you: the inquiries that lead to programmes that work read very differently from the ones that lead to a well-reviewed workshop and an unchanged organisation. The difference is visible before any contract is signed. It shows in what the inquiry asks for, in who signs it, and in whether the organisation understands what it is actually buying.

The trap: buying an event instead of a change

The default way to "do AI training" is to book a workshop. Someone credible comes in, demonstrates impressive things, the feedback forms are positive, and the initiative can be reported upwards as delivered.

On its own, it changes very little. I have written before about what makes AI training actually stick, and the evidence is consistent: behaviour change requires repetition, practice and accountability. A single event supplies none of those. People return to their desks, meet the first point of friction, and revert to how they worked before.

The cost of getting this wrong is not the workshop fee. It is visible in the licence data: organisations buy AI seats in bulk and then watch them sit largely unused. That gap is not a technology problem, and it is not solved by a better demo. It is an adoption problem, and adoption is precisely the thing a one-day format cannot deliver.

The question is not "who can run a great workshop?" It is "who can get my people genuinely working differently a year from now?" Those are different suppliers, and they answer an inquiry differently.

What you are actually buying

Reframe the purchase and the shortlist changes. You are not buying days of a trainer's time. You are buying a measurable shift in how a group of people work, one that holds after the trainer has gone home.

That means the thing to evaluate is not the quality of the demonstration. It is whether the provider has a credible theory of how change happens in an organisation like yours. A partner who understands this will ask about your workflows, your current adoption and your leadership before they talk about their content.

AI adoption is a leadership programme, not a procurement exercise

This is the part most inquiries miss entirely, so I will give it the space it deserves.

Every AI adoption programme carries an unspoken question into the room: are they doing this to replace us? Your people may never say it aloud. They will answer it for themselves anyway, and they will answer it by watching leadership, not by reading the internal comms.

If leaders are absent from the programme, if they talk about AI as an efficiency measure and never use the tools themselves, the workforce concludes that the honest answer is yes. What follows is rarely open refusal. It is quiet resistance: sessions attended but not applied, tools tried once and abandoned, plausible reasons why AI "doesn't really work for our team". A programme can be sabotaged politely, and underinvested leadership is the most common cause I see.

The organisations where training produces real change share a pattern, and it is set before any trainer arrives:

1 · Invested

The programme has a named senior sponsor

Someone whose reputation is attached to the outcome, not a delegated budget line. When adoption stalls, there is a person with authority who cares.

2 · Present

Senior people sit in the sessions

With their teams, doing the exercises. Not opening the day with a video message and leaving. Presence is the message.

3 · Leading by example

Leaders use the tools visibly, including imperfectly

And they answer the replacement question directly, in words and in behaviour: AI is here to raise the level of the work, not to thin the room.

The hard truth: no training partner, however good, can outweigh what your own leadership signals. If leadership is not prepared to be invested, present and visible, fix that before you send a single inquiry.

A serious partner will raise this in the first conversation; some of my most useful early engagements with clients are precisely this discussion, before any curriculum is designed. A supplier who is happy to take the booking without asking about leadership involvement is telling you which kind of supplier they are.

Put your most senior people in the room

The second pattern follows from the first. When an organisation treats AI training as routine procurement, the selection gets delegated downwards: a junior team member or an intern is asked to gather proposals, compare prices and shortlist three options. For commodity training, that is fine. For AI adoption, it is a mistake that shapes everything downstream.

Providers calibrate to the room. A junior employee collecting quotes receives brochure answers and a price list. A senior leader asking hard questions about adoption receives the provider's real thinking. You learn more in one direct conversation than in any written proposal.

This is not really a training purchase. It touches how your teams will work, what your data governance permits, and what your organisation is signalling to its people about their future. Those are leadership questions, and they cannot be evaluated from a comparison spreadsheet.

Conviction is built in the selection room. A leader who joined the conversations can stand in front of the organisation and say why this partner and why now. A leader handed a recommendation cannot, and their teams can tell.

The rule: the most senior person who will own the outcome joins the serious conversations, and the inquiry says who that person is. It tells every capable provider you are buying a change, not comparing day rates.
Before you brief anyone

Know where your teams actually stand

The AI Maturity Assessment scores a team across five dimensions — Awareness, Tool Adoption, Workflow Integration, Measurement and Culture — in 5 minutes. It gives you a shared baseline to put in your inquiry, so proposals target the real gap instead of guessing.

Take the Assessment →

What to look for in a long-term partner

Six things separate a partner from a performer. Weight them however fits your organisation, but do not skip them.

1. Bespoke, not off the shelf

Generic AI training teaches tools in the abstract. It shows a marketer's use case to a finance team and wonders why nobody applied it. A partner builds the content around your teams' actual tasks, tools and language, so people see their own job on the screen.

2. Behaviour change, not awareness

Ask how they design for what happens between and after sessions. If the entire offer is a single day with no practice loop, no follow-up and no accountability structure, you are buying awareness. Awareness is inexpensive, and it fades on schedule.

3. A series, not a single event

The most effective corporate AI programmes run as a structured series, typically three to five sessions spaced over several weeks, so people practise on real work between sessions and bring their obstacles back into the room. One session is a launch, not a programme.

4. A practitioner who uses the tools

AI changes monthly. A trainer who last used the tools in earnest six months ago is teaching history. Ask what they use in their own work and how recently. The credible partners run the experiment themselves before they teach the conclusion.

5. Fluent in your risk and data context

If you operate in financial services, legal, healthcare or the public sector, safe use is not a footnote. A serious partner can talk precisely about what should and should not go into an AI tool, and how to build confident use inside your compliance boundary rather than around it.

6. Accountable for what happens after the room empties

This is the clearest test. Ask directly: what does success look like in six months, and how will we know? A one-day supplier has no answer, because their responsibility ends when the day does. A partner will have one, because adoption is the entire point of their work.

What your inquiry should include

The quality of the proposals you receive is set by the quality of the brief you send. A vague inquiry — "we'd like some AI training for around 40 people, please send pricing" — produces vague, off-the-shelf responses, because you have given the provider nothing to design against. Thirty minutes on the brief buys you proposals you can actually compare.

Your AI training inquiry checklist

Copy this into your brief. The more of it you answer, the sharper and more comparable the proposals you get back.

  1. Who the training is forTeam or teams, number of people, roles, seniority, and how confident they actually are with AI today.
  2. Which AI tools you use — or are choosing betweenClaude, Microsoft Copilot, ChatGPT, Gemini, a mix, or "still deciding". If you have licences sitting unused, say so.
  3. The business outcome you actually wantNot "AI awareness". Something concrete you could measure: cut reporting time, lift weekly active use, get the team drafting confidently with your chosen tool.
  4. Your current stateHas this team been trained before? What is adoption like now? Was there a past programme that did not stick, and why do you think that was?
  5. Constraints and data rulesSector, regulation, and what can and cannot go into an AI tool. This shapes the whole design, so state it early.
  6. Format and logisticsIn person or online, location, languages, preferred dates, number of sessions and group sizes.
  7. Timeline and budget rangeWhen you want to start and a realistic range. It lets providers propose something honest rather than guessing at your ceiling.
  8. How you will measure successWhat "it worked" looks like in three to six months, and what data you can see: usage figures, confidence measures, time saved.
  9. One-off or ongoing partnershipBe explicit if you want adoption support beyond the sessions. It changes who takes your inquiry seriously.
  10. Who decides, and by whenName the senior people in the process and the decision date. Serious buyers get serious answers.

The questions that separate a partner from a performer

Once you have a shortlist, a direct conversation tells you most of what you need to know. Ask these, and listen for whether the answer is about your organisation or about their product:

  • Do you tailor the content to our tools and workflows, or is it the same programme for everyone?
  • How do you build behaviour change beyond a single session?
  • What role do you expect our leadership to play, before and after the sessions?
  • Do you use these tools in your own work, and how recently?
  • How do you handle our sector's data and compliance rules?
  • What happens after the training: do you support adoption, and how do you measure it?
  • What does success look like in six months, and how will we know we got there?

The third question earns its place. A provider who has no expectations of your leadership has not thought seriously about why programmes fail.

Red flags worth acting on

Some signals are reliable enough to end a conversation early:

  • A fixed one-day package quoted before anyone has asked about your teams.
  • Slides about AI instead of live tools.
  • A trainer who cannot say when they last used the tools in their own work.
  • No view on leadership's role, and no plan for what happens after the room empties.
  • A pitch about the technology that never once mentions your people's actual work.

Each of these is a supplier describing their own limits. Believe them.

Choose for the year, not the afternoon

AI is not a topic your organisation covers once and moves past. The tools change quarterly, your teams' needs change with them, and the organisations pulling ahead treat adoption as an ongoing capability with sustained leadership attention.

The test: will this partner still be useful to us in a year? The impressive one-day performance rarely passes it. The partner who asked about your leadership and your workflows before quoting a fee usually does.

Put your senior people in the room, send the inquiry that deserves a serious answer, and choose the supplier who is planning for your next twelve months rather than their next booking.