Question about Usage Guidlines

We are a tech company in Cornwall, UK who are working in helping people on low incomes access services. Our client is not the government but the user. I have been in this field for years and have done user testing with homeless people and people on the edge of poverty. I saw AI as a way for these people to access services and knowledge to increase their opportunities. We got funding to do this and see OpenAI as a way of creating solutions. For instance, I noticed that homeless people were using Messenger with the local library WIFI. So we want to create a bot that they communicate with, which a friendly tone and talks to them in common English but then fills out a lot of the housing applications for them. It would be a game-changer for these people. I want users to access information about benefits in friendly dialogue using Whatsapp.

I see that there are strict guidelines from OpenAI around this. We have a fast-growing team and especially with the fine-tuning launch, we can build these solutions. We do not represent the local government or any authorities but only the user so the AI will not be making any decisions on their housing or other benefits but giving them the ability to access these services. Will we be able to do this with OpenAI. I really hope we can be allowed to help these people with this API. Imagine what a homeless person with a Data Scientist could do?

Any advice before we move down this path would be appreciated.

1 Like

Hey @wo,

Very cool idea! And appreciate you checking in regarding potential safety guidelines. Since the app won’t be making housing decisions but rather is making the information more accessible I don’t anticipate any issues with this idea - seems like it would be very useful!

We do have a pre-launch review process as described here, during which we can work with you on any risk mitigation features that might be needed. I’ve flagged your idea to the review team, looking forward to seeing your application :slight_smile:

Good luck!