One of the most popular guides I shared in 2025 focused on a feature many teachers still overlook when using ChatGPT: Custom GPTs. Even now, I am surprised by how often I see educators rely exclusively on the general chat window, repeating the same context again and again.
If you are already using ChatGPT for teaching, creating a Custom GPT for your class should feel like a natural next step, not an advanced or optional move.
Moving beyond the general chat window
The default ChatGPT interface treats every conversation as a fresh start. Each time you open a new chat, you need to restate your role, your subject, your students’ level, and your expectations. Over time, this repetition becomes tiring and inefficient.
A Custom GPT changes that workflow completely. Instead of starting from zero, you create a dedicated teaching assistant that already knows your instructional context. You define its role, tone, and focus once, then let it work within those boundaries.
What you can upload to a Custom GPT
The real strength of a Custom GPT lies in the materials you give it. Teachers can upload:
• course syllabi
• lesson plans
• assessment rubrics
• curriculum standards
• slide decks and notes
• sample assignments and anonymized student work
Once these resources are in place, the GPT responds using your materials as its reference point. Lesson ideas align with your curriculum. Feedback reflects your rubrics. Explanations match the language level of your students.
This alone saves a significant amount of time compared to reintroducing context in every new chat.
Why this matters for teaching practice?
Custom GPTs are not about automation for its own sake. Their value comes from consistency and alignment.
When your AI assistant works inside a defined instructional framework, it produces outputs that are closer to what you would design yourself. That consistency helps with:
• Maintaining a stable tone across materials
• Supporting differentiation without redesigning tasks each time
• Drafting assessments that match learning outcomes
• Generating prompts and questions that reflect your teaching philosophy
A short guide for teachers
To make this process easier, I created a short teacher guide that walks through how to build a Custom GPT step by step. It explains how to define the GPT’s role, what types of materials to upload, and what ethical boundaries to keep in place, especially around student data and over-reliance on AI outputs
The guide also includes a concrete classroom example, showing how a secondary teacher might design a subject-specific GPT aligned with local curriculum standards.
Projects are now part of the same conversation
It is also worth noting that Custom GPTs are no longer the only way to work with contained context. The Projects feature, now available across ChatGPT, Gemini, and Claude, serves a very similar purpose.
Projects allow you to create a dedicated workspace where you upload materials, chat within that context, and keep everything organized in one place. The interface is different, but the underlying idea is the same: AI works best when it is given boundaries, resources, and a clear purpose.
The bigger shift
At a deeper level, this is about changing how we think about AI in teaching. Treating AI as a blank slate leads to shallow, generic outputs. Treating it as a contextual assistant leads to more meaningful support.
If you are a teacher still relying only on the general chat window, you are likely spending more time than necessary and getting less value in return. Custom GPTs and Projects offer a more professional, more sustainable way to integrate AI into everyday teaching work.
The technology is already there. The question is how intentionally we choose to use it.
