Google made a minor but significant update to their privacy policy over the weekend, effectively claiming ownership of anything you share publicly online for use in training their AI models. The section of their privacy policy’s previous wording said that public data would be utilized for business reasons, research, and to improve Google Translate services.
As of July 1st, 2023, the new adjustment is already in effect. Given the scale and length of Google accounts (consider how long some people have had Gmail and YouTube accounts), this update now formally incorporates a massive amount of public interaction data spanning decades. What is unclear is whether those who have vowed to “de-Googling” their online lives will be ensnared in the dragnet of Google’s data scraping regardless of whether they consent to this policy change, or if merely having any contact with Google throughout the years will enough. With other large language training models, such as OpenAI’s ChatGPT, large-scale public scraping has previously occurred without user agreement. Ideally, though, this move should only effect those who have current accounts with multiple Google services.
One critical item to note is that Google makes no mention of using private data, and anything supplied with Google appears to be secure from being swallowed into the AI system. For the time being.
