Google is already testing some of its AI creations, including its Magic Compose in the Google Messages app. The feature can help refine your composition, but it is important to note that it sends your data to Google’s own servers in order to generate suggestions.
Google disclosed this on its support page for the feature. “Up to 20 previous messages, including emojis, reactions, and URLs, are sent to Google’s servers and only used to make suggestions relevant to your conversation. Messages with attachments, voice messages, and images aren’t sent to Google servers, but image captions and voice transcriptions may be sent.”
Despite this, the search giant clarified that it doesn’t store the data, nor will it use it for training its machine learning models. Specifically, the company stressed that the messages sent to the servers would be discarded after the feature successfully generated the suggestions you need.
The feature will work even when users have end-to-end encryption (E2EE), which prevents any company from reading your message. However, on its support page, Google implies that although the feature will send the data to its servers, using RCS with E2EE will still prevent the company from reading the users’ messages. Also, it noted that the sending of data wouldn’t happen “when you are not using Magic Compose.”
Google’s disclosure of this matter reflects tech companies’ cautious moves in promoting their new AI creations to the public, which remains partially doubtful about the tech. Other companies are also trying to be more careful in embracing AI, with some ordering AI bans due to data leak fears. This, however, won’t stop Google from continuously investing in AI as it tries to steal back the spotlight from its rival company Microsoft. To recall, Microsoft’s ChatGPT-powered Bing and Edge reportedly caused Google to be on “Code Red.” This pushed the company to take drastic moves to expedite its AI projects, which it unveiled during its I/O 2023 event. And now, Google is testing a whole bunch of them in hopes of catching up with Microsoft’s AI-filled ecosystem.