Home Tech/AIGoogle refutes ‘deceptive’ claims about Gmail utilizing your emails to train AI

Google refutes ‘deceptive’ claims about Gmail utilizing your emails to train AI

by admin
0 comments
 
acastro_STK459_09
Jay Peters
is a senior journalist focusing on technology, gaming, and several other topics. He became part of The Verge team in 2019 after spending nearly two years at Techmeme.

Google is countering viral social media messages and articles such as this one from Malwarebytes, asserting that Google has modified its policy to allow the use of Gmail messages and attachments for AI model training, and the sole method to opt out is by disabling features labeled as “smart” like spell check.

However, Google representative Jenny Thomson informs The Verge that “these accounts are deceptive – we have not altered any user settings, Gmail Smart Features have been in place for many years, and we do not use the content of your emails for training our Gemini AI model.”

It’s advisable to verify your settings nonetheless, considering one Verge employee mentioned they had opted out of certain Smart Features, yet found themselves automatically re-enrolled. In January, Google refined its smart feature settings enabling users to deactivate features for Google Workspace and other Google services (like Maps and Wallet) independently.

Apart from functionalities like spell checking, enabling Gmail’s smart features allows for options like tracking shipments or seamlessly integrating flights from Gmail into your calendar. Activating this setting in Workspace indicates that “you consent to let Google Workspace utilize your Workspace content and activity to customize your experience across Workspace,” as stated on the settings page, yet Google clarifies that this doesn’t imply transferring your email content for AI training use.

You may also like

Leave a Comment