Google has unveiled Gemini, a new artificial intelligence chatbot specifically developed for children under 13. The tool will debut in the U.S. and Canada within the next week, with an Australian rollout planned later this year. Accessible exclusively through Google’s family-linked accounts, the chatbot gives parents oversight—but not without potential pitfalls.

During setup, parents must provide their child’s name and birthdate, sparking immediate questions about data privacy. Google asserts this information won’t train AI systems, yet the chatbot activates by default—requiring manual deactivation by guardians.

Children can interact with Gemini through text prompts or image generation requests. While designed to encourage engagement, the system isn’t foolproof. Google openly acknowledges occasional inaccuracies in responses, urging families to critically assess outputs. Unlike conventional search engines that retrieve existing information, AI tools synthesize new content—a distinction that could confuse young users navigating digital landscapes.

The tech giant has implemented safeguards against inappropriate content generation, but these filters risk unintended consequences. Overzealous restrictions might block legitimate educational material alongside truly harmful content.

Australia’s eSafety Commission has already sounded alarms about AI companions potentially exposing children to harmful advice or distorted realities. Commissioner Julie Inman Grant emphasizes the particular vulnerability of developing minds: "Young children often lack the critical thinking skills to identify misinformation—especially when it comes from what appears to be an authoritative digital source."

This launch coincides with Australia’s impending social media ban for under-16s starting December 2025. As governments tighten regulations, experts stress dual responsibilities: tech companies must prioritize child safety in design, while parents need to actively guide digital literacy.

Key Points

1. Google’s Gemini AI chatbot targets children under 13 via family-linked accounts, launching first in North America
2. Privacy concerns emerge as parents must submit children’s personal data during setup
3. Default activation requires parental intervention to restrict access
4. Australian regulators warn about risks including misinformation and harmful content exposure
5. Safety filters may inadvertently limit access to appropriate educational resources