ChatGPT Unleashed: Now Messaging Friends Directly via Apple
OpenAI has launched an Apple Messages plug-in for ChatGPT, allowing users to sort, analyze, and draft messages directly through the chatbot. While offering enhanced conversation management, the new feature raises privacy concerns regarding data retention and access, despite OpenAI's assurances of local processing and explicit user consent.
OpenAI has introduced a new Apple Messages plug-in for ChatGPT, enabling users to integrate their Messages inbox directly with the chatbot. This new functionality offers a range of benefits, allowing users to efficiently manage their digital conversations. With the plug-in, individuals can sort, analyze, or edit their messages directly from within ChatGPT. Beyond personal use, the plug-in is also compatible with OpenAI's Codex and ChatGPT Work platforms, extending its utility to professional applications.
A promotional demonstration of the plug-in showcased a user requesting ChatGPT to suggest follow-up messages based on prior day's communications. The capabilities extend further, including the ability to ask ChatGPT to delete messages, draft and send new messages on a user's behalf, or conduct searches for specific information buried within extensive message histories.
However, the introduction of this feature has sparked discussions regarding privacy. While OpenAI has outlined certain permissions protections, the overarching privacy framework remains somewhat ambiguous. In response to inquiries, OpenAI clarified that ChatGPT does not construct a comprehensive index of a user’s messages. The company also stated that the Messages plug-in operates locally on the user's device, and for ChatGPT to access messages, a user must explicitly request it. To fully enable the plug-in, users are required to grant Full Disk Access, which provides the application with the necessary permissions to read and write files on their device.
Despite these explanations, the complete details concerning what type of information, if any, is retained by ChatGPT are not entirely clear. OpenAI has been contacted for additional clarification on this matter. Furthermore, when it comes to sending messages, OpenAI strongly advises users to actively monitor ChatGPT's actions and cautions against enabling persistent approval. The company highlights that persistent approval removes the user's final opportunity to review a message before ChatGPT dispatches it, emphasizing the importance of user oversight in this process.