Choosing a platform: an important update
The course framework refers to building a “custom GPT”. As of this writing, OpenAI’s help centre states that custom GPTs are scheduled to retire on 11 December 2026 (11 February 2027 for some Enterprise workspaces with an approved deferral), that the change affects all ChatGPT plans, and that creators can migrate a GPT to a plugin. On migration the GPT’s instructions become a “skill” inside the plugin and its knowledge files are copied as reference files; custom actions do not transfer. OpenAI’s help pages also describe the creation of new GPTs as limited to Business, Enterprise and Edu workspaces with permission, with creation planned to end before retirement.
What this means for you: do not invest in a new custom GPT. The skills in this module (scoping, writing instructions, preparing knowledge, testing) transfer directly to whichever platform you use. If your organisation already relies on a custom GPT, read the FAQ and plan its migration and re- testing before the retirement date.
The options below all follow the instructions-plus-knowledge pattern. Availability and limits depend on your plan and change often, so check the linked documentation.
| Option | What it is | Good for |
|---|---|---|
| ChatGPT Projects | A workspace holding custom instructions, uploaded files and related chats. OpenAI states Projects are available on all plans, with file limits that vary by plan, and can be shared. Documentation | The simplest way to complete this module’s assignment in ChatGPT. |
| ChatGPT plugins and skills | OpenAI’s successor to custom GPTs. A plugin can bundle skills (reusable instructions) and connected apps, and is created by chatting with the Plugin Creator. Access varies by plan and workspace. Documentation | Teams on ChatGPT business plans who want a shareable, maintained assistant. |
| Claude Projects | A project with instructions and a knowledge base of uploaded documents, in Anthropic’s Claude. | An equivalent build in Claude. |
| Google Gemini Gems | Saved custom versions of Gemini with instructions and files. | Organisations using Google Workspace. |
| Microsoft Copilot agents | Agents built in Microsoft 365 Copilot or Copilot Studio, which can use documents in SharePoint and appear in Teams. | Organisations using Microsoft 365, where employees already work in Teams. |
| Zapier Chatbots | A chatbot builder inside Zapier that can use knowledge sources and trigger automations. | Linking an assistant to the automations you build in Module 5. |
| Assistants built into HR platforms | Many HR service and HRIS products include their own employee assistants. | Production use at scale, with HR-specific permissions. |
How to choose
- Where are your employees already? An assistant inside the chat tool people use daily gets used.
- Is the platform approved for internal documents? Check the data-protection terms and administrative controls.
- Who can access it, and can that be limited? A policy assistant for employees should not be public.
- Can you see what people ask? You need this for the feedback loop, and you need to tell users that conversations may be reviewed.
- What does it cost per user?
Jot down key ideas in your own words. Saves automatically.