Preserve a no-code, shareable course chatbot for students after custom GPT retirement

ChatGPT Feature requests
1

Please preserve a no-code way for educators to create and share course-specific chatbots with students. A teacher can currently configure a GPT with course instructions, reference materials, and conversation starters, then give students a link. This lets students use a deliberately designed learning experience without each student having to build or configure it.

The proposed Plugin migration does not carry over a GPT’s sharing settings or give existing users access to the replacement. A migrated personal plugin starts private and public availability requires a separate submission process. Building a Site or API app is possible, but requires development and maintenance that many individual instructors cannot take on. Those options do not replace the ease of publishing a course GPT.

Please provide a teacher-manageable replacement that supports a shareable class link, course-specific instructions and materials, straightforward updates during the term, and student access without each student installing or configuring a workflow. A class link should not require submitting every course bot for public directory review. Until a comparable path exists, please preserve existing course GPT links or provide a transition that keeps students’ access intact.

OpenAI’s retirement FAQ says migrated plugins do not inherit sharing settings: https://help.openai.com/en/articles/20001519-custom-gpt-retirement-and-migration-faq

2

The syllabus chatbot is an easy-to-understand example of a broader need. I teach instructors through the University of Florida’s Center for Teaching Excellence, as well as business professionals, to create purpose-built AI experiences. These include guided study, tutoring, role-play, simulations, and professional practice, not only course Q&A.

The practical test for a replacement is whether someone with no technical background can create one during a workshop: set its instructions and scenario rules, upload source files, test it, and share a public link to the specific experience. They should be able to replace a file or revise the instructions later without rebuilding software or disrupting access for the people using it.

Plugins may provide some of those capabilities technically. The part I hope OpenAI preserves is the complete no-code path from creator to audience, including a workable transition for existing GPTs and their users.