By Anna Rascouët-Paz
Publication Date: 2025-11-21 23:27:00
- In November 2025, a rumor circulated that Google’s latest update had automatically opted in all users to allow its generative artificial intelligence (AI) program to train on private communications such as emails and chats.
- While was true that Google’s AI and generative AI accessed users’ private content, it had been the case for a while. So-called “smart features” across Gmail and the Workspace platform required AI access to users’ private content. Snopes found tutorials posted as early as 2020 to help users disable smart features.
- A spokesperson for Google said in an email statement that it was “misleading” to say the company is training its generative AI model Gemini on users’ emails. Further, the person said Google did not change people’s settings. We asked Google for more details and will udpate this report should we hear back.
- In 2023, Google said it expanded its smart features with generative AI, which required…
