Google has issued a privacy warning to all Android and iPhone users concerning potential security threats associated with artificial intelligence in smartphone apps. The warning, published on Google's Gemini App Privacy Hub blog, advises customers to refrain from divulging confidential information during conversations on Gemini apps.
Gemini apps, described as akin to a highly advanced Google Assistant, are the focal point of concern. Google emphasized the importance of avoiding the disclosure of personal data, cautioning users against inputting any information they wouldn't want reviewed or utilized by Google for product enhancement.
Please don't enter confidential information in your conversations or any data you wouldn't want a reviewer to see or Google to use to improve our products, services, and machine-learning technologies, emphasizes Google's Gemini App Privacy Hub blog.
The blog elaborated on the retention policies, stating that conversations reviewed by human evaluators, along with associated data such as language, device type, and location information, are stored separately and remain accessible for up to three years, even after deletion from the Gemini Apps activity.
Furthermore, Google clarified that conversations, including potentially sensitive information, are retained for up to 72 hours in a user's account, even when Gemini Apps Activity is disabled. This period facilitates service provision and feedback processing, though such activity is not reflected in the Gemini Apps Activity log.
Notably, Gemini can inadvertently activate through voice commands resembling its wake phrase, such as "Hey Google." In such instances, Google assures users that unintended activations will be treated in accordance with their settings.