AI sent message without consent? Controversy swirls after user's claim

A user of Google’s artificial intelligence service Gemini has claimed the system sent a text message to an acquaintance without consent, raising concerns over unintended actions by AI systems.

The user, who identified as a Gemini user, shared the experience on an online community forum on Friday. According to the post, the user was conversing with the AI while discussing a hypothetical scenario involving smuggling into China when a message generated by the system was sent to a colleague without the user’s intention.

The text message was reportedly sent at dawn to a colleague the user was not close to, leaving the user confused.

Following the post, other users reported similar experiences, particularly among those using the Android operating system. Some said text messages had been sent unintentionally during conversations with AI services, while others claimed the AI had even initiated phone calls.

Gemini currently supports sending text messages and making phone calls on Android devices. When a user requests that a message be sent to a designated phone number, the system sends the message after confirming it is connected to Google Assistant. The feature is not available for iPhone, but it could be expanded in the future, as Apple has hinted it plans to integrate Gemini into its virtual assistant Siri.

Regarding the recent case, Google said the user may have responded “yes” when the AI asked whether to send the message. However, concerns persist that sensitive content could be delivered to unintended recipients if users unknowingly approve actions during conversations with AI systems.