Setting Up a Teacher Workspace

From Personal Tool to Team Asset

Until now, ChatGPT has been your personal tool: your conversations, your prompts, your organization. This lesson scales up. We’ll cover how to configure a structured workspace for an entire teaching team—a department, a school, or a network of schools. This shift isn’t just technical. What works for a solo user, like keeping your best prompts in a text file, collapses when thirty colleagues need to share the same resources and adhere to a single data policy.

The Architecture of an Educational Workspace

A well-organized ChatGPT Edu workspace relies on three nested levels. The organization corresponds to the school or network and is managed by a primary administrator. Groups divide this organization into disciplinary departments, project teams, or grade levels. Finally, members are the teachers and, in higher education, the students, each holding distinct roles and permissions. This hierarchy isn’t decorative; it’s what lets you enable a feature for science teachers without exposing it to sixth-grade classes.

Each level carries a responsibility that someone must explicitly own. The table below summarizes who does what in a typical deployment.

Four Configuration Steps

Step 1: Define Your Data Policy

Do nothing else before this. It’s far easier to set a rule on an empty workspace than to enforce it across a thousand existing conversations. Decide on conversation retention periods, then determine which features you’ll enable or disable—memory, link sharing, projects, available models. New eligible Enterprise or Edu customers can also choose to store data and process requests in Europe. Have the Data Protection Officer validate the entire setup: legal basis, user notification, and impact assessment if required. A school that launches its workspace in September without settling these four questions will discover them the moment a parent files a complaint.

Step 2: Create Your Group Structure

Groups should reflect how your school actually works, not a theoretical org chart. Departmental splits—Humanities, Sciences, Languages—work best where disciplinary teams are strong. Grade-level splits, from middle school through high school to vocational tracks, make sense when needs differ primarily by student age. Cross-functional project groups, whether for career pathways, interdisciplinary units, or special initiatives, temporarily bring together colleagues from different backgrounds. You can combine all three approaches, provided you accept that a teacher may belong to multiple groups.

Step 3: Set Permissions by Role

In workspace settings, the permissions and roles section controls, role by role, which models, projects, memory features, link sharing, or ChatGPT Work access are available. A new Enterprise or Edu workspace starts with a selection of enabled applications, and new plugins and apps are generally disabled by default. Review the list in the admin console under Plugins, and leave open only what a specific group needs. A language department producing visual aids has different requirements than a math team. Leaving everything open everywhere is the same as making no decision at all.

Step 4: Prepare Shared Resources

This step separates a workspace where everyone starts from scratch from one where the team builds on shared assets. Prepare shared projects for common uses: an Official Curricula project with current standards, an Administrative Writing project with letter and report templates, an Assessment project with your multiple-choice guidelines and rubric criteria, and an Inclusion project with adaptation principles for students with special needs—PAP, PPS, PPRE plans—never including any student data. Your recurring instructions can also become a plugin, grouping skills—reusable instructions and procedures—and, if needed, applications. Someone with permission publishes this to the workspace plugin directory. Shared GPTs, long used for this role, retire on December 11, 2026; their published instructions and knowledge can migrate to a plugin. A teacher joining mid-year thus finds an environment already calibrated to your school’s practices.

Security and Monitoring

Certain data must never enter a conversation, regardless of contractual guarantees. Real student names, medical or social data, graded results with names, photos, and identity documents stay out of scope. ChatGPT Edu provides serious protections, but data minimization remains the golden rule—and the only one that protects you if an incident occurs at the provider’s end.

On the monitoring side, the admin console for Enterprise and Edu workspaces displays usage—active users, messages, projects—broken down by group. This helps you spot a group using nothing, another blocked by a disabled feature, and adjust permissions to match actual consumption. For compliance needs, a dedicated platform also lets the organization export conversation logs and content to its own archiving and auditing tools. Members must know this.

Your Turn

Design your workspace.

1. Sketch your ideal workspace on a single page: list the groups and explain why
2. List the data decisions requiring Data Protection Officer approval
3. For each role, note which features are enabled and which are disabled
4. Choose the top three shared projects to prepare first
5. Present this written proposal to your leadership team with a timeline

Official Sources: ChatGPT Edu at OpenAI · Managing workspace settings in ChatGPT Enterprise · Workspace analytics for ChatGPT Enterprise and Edu · OpenAI Compliance Platform for Enterprise and Edu customers · Data residency and inference residency for ChatGPT · Projects in ChatGPT · Plugins in ChatGPT · Admin controls, security, and compliance for plugins and apps · Custom GPT retirement and migration FAQ (OpenAI Help Center) · Data controller: how to implement AI systems in education? (CNIL, France’s data protection authority)

Key Takeaways

• A structured workspace—organization, groups, members—is the foundation of a successful deployment
• Configure your data policy before inviting users
• School-wide shared projects prevent every teacher from reinventing the wheel
• Data minimization remains the rule, even with ChatGPT Edu