From what I know, the warnings don't trigger any automatic action from OpenAI, so it's an arbitrary amount. You'd probably have to do a lot of policy-breaking stuff before they disabled your account. I wouldn't worry about it.