We need to let users to try out openai api-calls within jupyter notebook like online code editor environment,
openai python API calls (hands-on tutorials for internal training)openai.api_key or if user tries to print it should print only encrypted value, within the online code editorHi JoshuaB! Thanks for bringing this question up!
What I usually do, when running notebooks and needing an API Key, is have a cell with this code:
from getpass import getpass
# type your API KEY, once asked
API_KEY = getpass('Enter API key: ')
So, you might be asked to copy/type your API key without anyone “seeing it” and now have the key in this variable API_KEY
Hope this answers your question and helps! 
Thanks @lesly.zerna for the reply.
our requirement is not just hiding openai.api_key while typing, during the session. since we need to provide access to the notebook itself (similar to the short courses), but, need to mask/encrypt when the user tries to print / access openai.api_key(limiting api_key usage only to that particular notebook)
Any pointers to how api_key access is being controlled, in notebooks shared with short courses will be a great help.
Hey Guys,
Yes I too think that this looks like a security concern of itself. I can really just access the OpenAI API Key that Deepliearning has by executing
openai.api_key
That looks like free API Calls to me. (duh I know you must have it set rate limits, but still…)
@lesly.zerna I’m sure engineers at deeplearning.ai are aware of this. How is the dev team explaining that this is not a security break?
Thanks for reporting! I’m checking that with the team! Which course in specific are you checking? Thank you!
