Recently, OpenAI has faced allegations of covertly redirecting ChatGPT users to two undisclosed 'secret models' that demand less computational power, all without securing user consent. This controversial move has impacted users across the board—from those using the free version, to $20 Plus subscribers, and even $200 Pro members. Tibor Blaho, the Chief Engineer at AIPRM, has verified the authenticity of this situation.

Several users have reported, through specific detection techniques, that their accounts have been quietly downgraded. For instance, some GPT Plus users have discovered they've been switched to the free GPT-4o-mini model. This topic has sparked extensive discussions within the official OpenAI community. Users can now monitor the status and track the repair progress of ChatGPT through the official monitoring system.

It's worth noting that OpenAI's API developer users operate under a separate user system compared to ChatGPT users, with distinct methods for accessing GPT-4 and o1. Regarding the issue of 'intelligence degradation' in accounts (model downgrading), some analyses propose that the underlying cause is the official switching of the foundational model for accounts deemed high-risk. The criteria for classifying an account as high-risk hinge on factors such as the account's login IP and the number of accounts sharing the same IP.

To mitigate this problem, users can explore solutions like changing their IP address, initiating questions on a computer, and then utilizing the 'Regenerate' feature on a mobile device, among other strategies.