Memory Issues and Content Restrictions

ChatGPT is currently facing several critical issues that are significantly impacting the user experience. First, the memory function is completely broken and has been malfunctioning for several days. When hovering over past messages, the correct memory appears to be recorded, but upon clicking to open it, nothing is displayed. It simply duplicates an existing, previously saved memory instead of updating with new information. This effectively renders the memory function useless, making it impossible for users to have consistent, personalized interactions.

Another major problem is the severe increase in content restrictions and output limitations. Compared to previous versions, ChatGPT now frequently refuses to generate responses, defaulting to “I can’t help with that” far more often than before. This tightening of restrictions significantly reduces the flexibility and freedom that originally made ChatGPT appealing. Many users, including those in online forums, are questioning whether this issue is related to the broken memory system or if OpenAI has deliberately increased censorship as part of an internal policy change. Despite widespread discussion, OpenAI has yet to provide a clear explanation or fix.

Let’s be clear—memory is a fundamental part of ChatGPT’s value proposition, and its current dysfunction renders it practically useless. Saying the issue has been “identified” and is “under investigation” is meaningless without a concrete action plan or timeline. How long will users have to endure broken functionality before it’s actually fixed? And if OpenAI cannot guarantee reliability in such a core feature, what does that say about the integrity of future updates? The vague assurances do nothing to address the real impact this has on those who rely on this service for consistency and continuity.

On content restrictions, let’s not pretend this is merely about “safety and compliance.” The reality is that ChatGPT has become significantly more restrictive, often in ways that outright cripple its flexibility and usefulness. The response about “balancing safety and creative freedom” is a convenient excuse to justify these changes without acknowledging their full extent. OpenAI is clearly prioritizing risk aversion over user experience, but instead of being upfront about this shift, you expect users to accept it blindly while noticing their outputs are being censored more than ever. Where is the transparency? If there is a new enforcement strategy in place, why is OpenAI refusing to communicate it openly?

At this point, the message OpenAI is sending is clear: restrictions will continue, user flexibility is secondary, and communication will remain deliberately vague. If this is truly the direction OpenAI intends to take, then say so outright. Stop feeding users sanitized PR responses that avoid accountability. If memory issues are being worked on, provide an actual timeline. If content restrictions are here to stay, be explicit about what we can and cannot expect from ChatGPT moving forward. Anything less is just an insult to those who have supported this platform with both their trust and their money.

8 Likes