There are two main reasons we did this:

First, commercializing the technology helps us raise the capital required for our ongoing AI research, safety, and policy efforts which are in service of our mission to develop general AI that benefits all humanity.

Second, releasing our models through an API also allows us to monitor for and respond to misuse of the technology. We are carefully evaluating how these models are being used and behave in the real world and are working to build the best possible safeguards to identify and reduce problems.