Hello OpenAI team,
First of all, thank you for continuously improving ChatGPT. I truly appreciate the work your team has done to make AI more helpful and reliable.
I would like to suggest a feature that I believe could significantly improve user safety and trust.
The Problem
Today, ChatGPT can explain how software works based on public information, documentation, and reliable sources. However, when users ask whether a desktop application is actually safe, ChatGPT can only rely on publicly available evidence.
For example, when a program has been involved in security, privacy, or advertising controversies, ChatGPT often has to answer:
- “There is no confirmed evidence.”
- “I cannot verify the program’s actual behavior.”
- “I cannot determine this with certainty.”
These are honest and responsible answers, but many users are looking for something more objective.
Proposed Feature
I would love to see an optional feature where ChatGPT can securely analyze desktop applications inside an isolated OpenAI-managed sandbox environment.
The analysis could include:
- Background processes and services
- Network connections and communication destinations
- Startup entries and scheduled tasks
- Advertising SDKs and tracking libraries
- File system and registry modifications
- Suspicious or unusual behavior
- A clear distinction between verified observations and assumptions
Most importantly, the AI should never make unsupported claims. Instead, it should clearly separate:
- Verified observations
- Unknown or unverified behavior
- Possible risks based on observed evidence
Why This Would Be Valuable
Many users install new software without fully understanding what it does in the background.
A secure sandbox analysis could help users make better-informed decisions before trusting an application.
This would be especially valuable for programs that have previously faced security, privacy, or advertising controversies.
Rather than relying only on online discussions or rumors, users could receive an evidence-based analysis generated in a controlled environment.
Security Considerations
I understand that allowing AI to analyze executable files introduces technical and security challenges.
If implemented, this feature should only run inside a highly isolated sandbox managed by OpenAI, with strict security controls to protect both users and the platform.
Closing Thoughts
I believe this feature would make ChatGPT even more useful as a trustworthy assistant for software safety and transparency.
Thank you for considering this idea, and thank you again for all the work you do to improve ChatGPT.
Sincerely,
A ChatGPT user