ChatGPT's macOS App Under Threat
OpenAI’s ChatGPT recently introduced a Mac computers desktop version to offer a seamless experience, but its release was delayed due to a security vulnerability in data handling.
Plain Text Storage Threatening macOS
Pedro Veito, a data engineerr and swift developer, mentioned on Threads that the app has a major privacy flaw. According to Veito, the reason behind this security issue is that ChatGPT handles that data in a different manner than Apple, by storing all the user conversations in plain text, making it easier for anyone to gain access to them.
OpenAI’s storing manner, the conversations are stored in a defined location, which is on the library and the application support on the official website. Therefore, other apps, processes or even malware that might be operating on the Mac computers desktop can have access to the data without the user knowing it.
In contrast, to protect users’ data, Apple uses the standard macOS sandbox system, which separates data for security purposes.
Standard Sandbox System
Apple’s standard sandbox system is a security feature that allows applications and their data to operate in a separate space, aiming to avoid its accessibility to other parts of the system without clear consent. It is worth mentioning that on iOS, all types of third-party apps are sandboxed, but on mac computer desktop sandboxing has become an option that users can choose to function or not since the introduction of OS X Lion in 2011.
Afterwards, the Cupertino-based company decided to add an extra layer of security with macOS Mojave, which requires apps to always ask for the user’s authorization to access data that isn’t sandboxed
For OpenAI’s privacy policies, it considers the data provided by users to be gathered as it helps the company improve the capabilities of the AI model. That’s why users should be very careful when sharing their data with the chatbot. Additionally, the absence of encryption in ChatGPT raises the risk.
AI company did not yet provide any response related to the sandboxing issue.
Final Thoughts
It is well-known that the partnership between these two giants created a sort of surprise, given that Apple on its mac computers desktop is very committed to protecting users’ privacy and OpenAI is doing its best to improve all issues related to this matter. However, will this recently discovered security vulnerability push the AI company to work harder? Or will it affect the partnership?
Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Intelligent Tech sections to stay informed and up-to-date with our daily articles.