OpenAI has rolled out a major update to ChatGPT’s memory system. It introduced a feature that allows the AI to automatically forget less relevant information while keeping important details “top of mind.” The new system brings ChatGPT closer to how human memory works — retaining what’s used frequently and quietly letting the rest fade — making it more efficient, adaptive, and user-friendly.
Previously, ChatGPT’s memory allowed users to store personal details to maintain context across conversations. However, managing these memories required manual effort and often led to “memory full” notifications. The new automatic memory prioritization changes that by letting ChatGPT decide which information remains relevant based on recency, frequency, and user interaction.
Key highlights of the update include:
- Automatic memory management: ChatGPT can now self-organize stored data, prioritizing frequently mentioned topics and de-emphasizing outdated ones.
- User transparency and control: Users can still browse, edit, re-prioritize, or delete stored memories through settings.
- Streamlined performance: The update eliminates manual maintenance, reducing clutter and improving contextual accuracy across sessions.
- Global rollout: The feature is being introduced to ChatGPT Plus and Pro users on the web starting October 15, 2025.
OpenAI says the enhancement represents a shift from a reactive to an adaptive AI model, giving users a smoother long-term experience. It also aligns with the company’s broader push toward more agentic and persistent AI assistants that learn over time while maintaining transparency and user control.
OpenAI’s latest upgrade signals a growing emphasis on personalized, context-aware AI systems that balance automation with privacy. For many, this “forgetting to remember” feature may be the most human-like improvement yet.
Source:
