This seems like a user issue, rather than a OpenAI breach. FTA:
> The majority of the dumped credentials were found within logs connected to multiple information stealer malware families.
So, at least for the moment, OpenAI's security is not in question.