OpenAI said it banned several ChatGPT accounts that generated fake online articles and comments, adding that the content gained little to no reach.