# Train-the-trainer course or your own in-house AI trainers: which does your company need? > A public train-the-trainer course trains one person to teach. In-house AI trainers keep AI use going across departments. How to choose, for Hong Kong HR teams. - URL: https://hyperfocusam.com/blog/train-the-trainer-course-or-in-house-ai-trainers/ - Language: en - Author: Sam Wong (黃力桐), https://hyperfocusam.com/ - 中文版: https://hyperfocusam.com/blog/train-the-trainer-course-or-in-house-ai-trainers-tc/ (Markdown: https://hyperfocusam.com/blog/train-the-trainer-course-or-in-house-ai-trainers-tc/index.md) - Published: 2026-10-31; revised 2026-10-31 --- Search "train the trainer 課程" in Hong Kong and you will mostly find courses for individuals: you enrol, you learn to teach, you leave with a certificate. Search "AI champion" and you find something similar: public classes where HR professionals build their own AI skills over a few mornings. Both are useful. Neither is the same as building a group of your own staff who will keep teaching AI inside your company after the outside trainer leaves. This is how I would answer an HR manager who asks me which one to buy. ## The short answer Choose a public train-the-trainer course if one person needs to become a trainer. Build your own in-house AI trainers if your company needs AI use to keep going across departments after the training ends. ## What a public train-the-trainer course gives you A public course trains individuals from different organisations together. You work on course design, exercises and facilitation, and in a good programme you keep materials you can adapt. I teach in one. [Adaptig's trainer programme](https://hyperfocusam.com/en/cases/adaptig-ai-trainer-cohorts/) runs live online: 12 sessions over six weeks per cohort, with participants in several time zones. Eight trainers graduated from the Hong Kong founding cohort, and graduates have gone on to run corporate and public workshops in Hong Kong. One of them won a global luxury brand as her own client, and [the client came back for more](https://hyperfocusam.com/blog/trainer-graduate-repeat-client/). That worked because she wanted to teach for a living. She chose the course, practised the craft, and went looking for clients herself. Public "AI champion" courses are something else again. The one I looked at in Hong Kong runs as three modules of three hours, for HR professionals: telling better stories with AI, solving problems with AI, and working with AI agents. That builds a more confident AI user. It does not set out to teach that person how to run a session for 20 colleagues, or what happens in your company in month two. ## What in-house AI trainers give you An in-house programme trains a small group of your own staff together, on your approved tools and your real work. The goal is not a certificate. The goal is that 90 days later, people across departments are still using AI and know who to ask. The difference shows up in the details. **The tools.** A public course has to teach tools that work for everyone in the room. Your company may allow only one. At [Garden](https://hyperfocusam.com/en/cases/garden-ai-pioneer-programme/), the IT directive allowed only Microsoft Copilot on the web and Perplexity, and the most confidential tier of data was barred from AI tools altogether. The training had to work inside those limits, so it did. **The work.** In-house trainers practise on tasks their colleagues actually do. When they stand up in front of their own department, the room already knows the task. They only have to show the AI part. **What happens after.** A public course ends when the cohort ends. An in-house programme should end with a question channel, a regular clinic, and trainers who have already run a real session. [HKCT](https://hyperfocusam.com/en/cases/hkct/) is the closest example I can show you: a single whole-college talk became a monthly AI learning community, with more than 200 members in its WhatsApp question group, and later monthly workshops were run by other trainers on the same programme. ## Side by side | | Public train-the-trainer course | Your own in-house AI trainers | |---|---|---| | Who learns | Individuals from different organisations | A small group of your staff, from across departments | | The goal | One person can teach AI | AI use keeps going after the training | | Tools | What suits a mixed room | Your approved tools only | | Practice material | General examples | Your own tasks, with safe data | | Ends with | A certificate | A session already run, a question channel, a check at day 90 | | Best when | Someone wants to teach as a career | The company needs AI to spread across departments | ## Four questions to ask before you choose **Who will be teaching a year from now?** If the answer is one specialist, a public course may be enough. If the answer is "someone in every department", you need a group, trained together. **Do your staff know what they are allowed to use?** If not, sort that out first. Neither option works while people are unsure. I wrote about [staff AI data rules in Hong Kong](https://hyperfocusam.com/blog/staff-ai-data-rules-hong-kong/). **Is there time in the calendar?** Champion roles that sit on top of a full-time job tend to fade; I wrote about [why AI champions give up within a few months](https://hyperfocusam.com/blog/why-ai-champions-give-up/). In July 2026, the Conference Board found that only 48% of workers agree their organisation gives them enough time during work hours to build AI skills. Whichever option you choose, agree the hours in writing. **How is AI training funded?** HKIHRM's 2025/26 training survey found that 82% of Hong Kong companies now let staff use generative AI at work, but only 6% have a dedicated AI training budget. If yours is funded project by project, a group of your own trainers is one way to keep the value after the project ends. Some employers may also recover part of the cost of advanced-technology staff training through NITTP; eligibility is decided by the NITTP Secretariat, not by me. See [government funding for corporate AI training](https://hyperfocusam.com/blog/government-funding-corporate-ai-training-hong-kong/). ## One more thing: the mainland certificate Some people searching "AI 訓練師課程" are looking for the mainland 人工智能訓練師 qualification. That is a vocational qualification under a national occupational skill standard, covering data collection and processing, data annotation and the operation of AI systems. Neither option on this page leads to it. ## Where I come in If you want to teach AI yourself, I point you to Adaptig's 6-week Trainer Certification. If your company wants its own staff to carry AI, I train your in-house AI trainers. Some companies start with one and add the other later. A certificate shows that one person can teach. Your own trainers show that the company can keep learning. --- **Not sure which one your company needs?** Tell me how many staff you have and which AI tools you're allowed to use, and I'll suggest a first step. [WhatsApp me](https://wa.me/85264315177?text=Hi%20Sam%2C%20I%20read%20your%20article%20comparing%20train-the-trainer%20courses%20and%20in-house%20AI%20trainers.%20We%20have%20about%20__%20staff%20and%20use%20__.) · [How the in-house AI trainer programme works](https://hyperfocusam.com/ai-train-the-trainer-hong-kong/) ## FAQ ### Is an AI champion the same as an in-house AI trainer? Not quite. An AI champion is usually an enthusiastic user whom colleagues ask for help. An in-house AI trainer has also learned to teach: they run sessions, prepare exercises on the team's own work and keep a question channel going. ### Can one employee attend a public course and then train everyone else? It can work if that person has the time and wants to teach. The risk is that one person carries the whole company on top of their own job. ### Is an AI train-the-trainer course the same as the mainland 人工智能訓練師 certificate? No. 人工智能訓練師 is a mainland vocational qualification covering data work and AI system operation. A train-the-trainer course is about teaching people to use AI at work. ## Sources - Adaptig trainer programme: [/en/cases/adaptig-ai-trainer-cohorts/](https://hyperfocusam.com/en/cases/adaptig-ai-trainer-cohorts/) (read 31 October 2026). - One of my trainer graduates won a global luxury brand: [/blog/trainer-graduate-repeat-client/](https://hyperfocusam.com/blog/trainer-graduate-repeat-client/) (read 31 October 2026). - Garden: department-by-department AI training: [/en/cases/garden-ai-pioneer-programme/](https://hyperfocusam.com/en/cases/garden-ai-pioneer-programme/) (read 31 October 2026). - HKCT: from a whole-college talk to a monthly AI learning community: [/en/cases/hkct/](https://hyperfocusam.com/en/cases/hkct/) (read 31 October 2026). - HKIHRM, 3-Module Series, AI Champion Programme: [hkihrm.org](https://www.hkihrm.org/product/3-module-series-ai-champion-programme/) (read 31 October 2026). - Human Resources Online on the HKIHRM 2025/2026 Training and Development Needs Survey, July 2026: [humanresourcesonline.net](https://www.humanresourcesonline.net/ai-tops-corporate-training-priorities-but-soft-skills-remain-essential-survey) (read 31 October 2026). - The Conference Board, press release, 28 July 2026: [conference-board.org](https://www.conference-board.org/press/ai-skilling) (read 31 October 2026). - National occupational skill standard for 人工智能訓練師 (2021 edition): [osta.org.cn](https://www.osta.org.cn/api/sys/downloadFile/decrypt?fileName=4eNsZIOv3WFNDfbnldJHeA/2024/4/29/6761815658ea468cbc0caa8974357703.pdf) (checked 20 October 2026, as cited on the train-the-trainer page). --- *Sam Wong (黃力桐) is a co-founder and Director of Academy at Adaptig. He designs AI training for teams and trains the trainers who deliver it.*