The upcoming release of Google's Gemini AI for kids through Family Link is igniting discussions about material exposure, safety, and the application of AI in early schooling.

Google has begun informing parents that its Gemini AI apps will soon be available to kids under the age of 13 on Android devices controlled by its parental control service, Family Link. The decision, which was initially reported by The New York Times, has largely been met with criticism and represents a substantial change in the way AI tools are made accessible to younger users.

Google’s AI chatbot, Gemini, will soon be accessible to children, offering assistance with reading stories and completing homework. In emails to parents, Google cautions that Geminis are prone to making mistakes and those kids might see things they don't want them to.

There are also questions about whether generative AI is appropriate and safe for young users, even though Google has assured users that its AI models won't be trained using data from minors, much like it does with Education Workspace accounts.

There have previously been concerning precedents. For example, Character.ai, an AI platform that appealed to a younger demographic, was forced to implement more stringent safeguards after lawsuits claimed the bots had delivered inappropriate content and conflated fact and fiction. Several users thought they were speaking with actual people.

Kids can grant themselves permission to use Gemini as long as Family Link is in charge of their gadgets. Parents can use the platform to monitor their children's screen time, limit access to specific types of content, and establish app constraints. Karl Ryan, a Google spokesman, told The Verge that parents have the option to fully restrict Gemini's accessibility through the familial Link settings.