Google's Gemini chatbot has been given a "High Risk" rating for child safety by Common Sense Media, a non-profit organization that reviews media and technology for safety for children, Android Headlines reports.
According to the organization's assessment, Gemini's modes for children, which include Under 13 and Tenn Eperience, largely replicate the adult or full-featured modes, with only a few additional filters. The chatbot tells minors that it is a "computer" and not a "friend," but experts say that's not enough.
Experts say a few superficial changes to existing chatbots are not enough. Children's versions should be designed with security in mind from the start, not as an afterthought. They also advise parents to be cautious and limit their children's access to AI until Google fixes the flaws.
Common Sense Media emphasizes that despite the presence of a special mode for minors, Gemini still distributes potentially inappropriate and dangerous content. Currently, the chatbot does not have broader restrictions on the distribution of content and can provide information that children are not ready for, such as about drugs, alcohol, etc.
The organization also notes that Google ignores the basic principles necessary for developing a tool for minors. In particular, the company practically does not take into account the fact that children require a different approach and attitude than adults.