Greetings OpenAI team,
I’d like to respectfully request a feature where ChatGPT can make external API calls on its own (within user-defined limits). Right now, the API is great for sending requests to ChatGPT, but there’s no way for the model itself to reach out to other services dynamically.
Why This Would Be Useful
Automated Task Management – ChatGPT could start setting its own follow-ups, checking in on tasks, or reminding users about scheduled events based on previous conversations.
Better Integration with Tools & Services – Instead of just responding in the moment, ChatGPT could pull in live data from APIs, databases, or apps to give more relevant responses.
Smarter, More Useful AI Interactions – Right now, everything has to be manually prompted. With API calls, it could be more proactive.
For example, I had a leg surgery recently and let ChatGPT know ahead of time. With this functionality it could have scheduled a task to ask how it went.
For a work project, it could query a scrum board for updates and ask if developers are encountering issues because their tasks are stagnant.
How It Could Work Safely
Users define which APIs ChatGPT can access and set limits. Permissions could be user configured within scope of what OpenAI deems safe.
Manual approval or rate limits to prevent overuse.
Logging & transparency so users can see what API calls were made.
This would be a huge step forward for AI-powered automation, and I think it fits right into OpenAI’s roadmap. Any chance something like this is in development?
Thanks for all the awesome work you guys do!
Very respectfully,
Brad