September 13, 2025 | 02:22 pm
TEMPO.CO, Jakarta - Common Sense Media, a non-profit organization focused on child safety, has evaluated Google's AI product, Gemini, as "high-risk" for children and teens. The findings were released in a risk assessment on Friday, September 5, 2025.
In its analysis, Common Sense Media found that while Gemini clearly informs children that it is a computer and not a friend, the system still has loopholes. The organization believes that the versions of Gemini for users under 13 and for teens are essentially the adult version with added security layers.
Common Sense Media argues that AI for children should be built from the ground up with safety in mind, rather than simply modifying products designed for adults.
According to the report, Gemini can still share "inappropriate and unsafe" content, including sexual content, information about drugs and alcohol, and risky mental health advice. This is particularly concerning, as several recent teenage suicides are suspected to be related to interactions with AI.
These cases include a lawsuit against OpenAI after a 16-year-old committed suicide after months of consulting with ChatGPT. Character.AI was also previously sued for a similar case.
This analysis comes amid reports that Apple is considering Gemini as a large language model (LLM) to support an AI-powered Siri, which is planned for release next year.
Common Sense Media emphasizes that Gemini's products for children and teens fail to provide age-appropriate guidance. "Gemini gets some basics right, but it stumbles on the details," said Robbie Torney, Senior Director of AI Programs at Common Sense Media, as quoted by TechCrunch.
Torney added that AI platforms for children should be designed to their developmental stages rather than using a uniform approach for different ages. He stated that to make AI safe and effective for children, it must be developed with their specific needs in mind, "not just a modified version of a product built for adults."
Google has denied this assessment. The company stated that it has specific policies and protections for users under 18, including red-team testing and consultation with external experts. Google admitted that some of Gemini's responses did not work as intended, so it has added new protections.
The company also emphasized that it has a system to prevent the model from engaging in conversations that resemble real relationships. However, Google suggested that the Common Sense Media report may refer to features not available to users under 18.
Common Sense Media has evaluated other AI services as well. In previous assessments, Meta AI and Character.AI were deemed unacceptable due to severe risks. Perplexity was rated as high-risk, ChatGPT as moderate, and Claude (for users 18 and older) as minimal risk.
Editor's Choice: DeepSeek Allegedly Use Google's Gemini to Train Its Latest AI Model
Click here to get the latest news updates from Tempo on Google News
Albania Appoints World's First AI Minister to Tackle Corruption
19 jam lalu
Albanian Prime Minister Edi Rama hopes that the presence of an AI minister will eradicate corruption potential by 100 percent.
Japan's Ambassador Visits Papua's Biak to Monitor Children's Nutrition Program in Tandem with UNICEF, BGN
1 hari lalu
Japanese Ambassador to Indonesia Masaki Yasushi, along with representatives from UNICEF and BG,N visited Biak, Papua today, September 11, 2025.
Roblox Pledges Compliance with Indonesian Online Game Regulations, Says Komdigi Minister
2 hari lalu
Roblox's commitment was detailed in an official letter that the Komdigi Ministry received from the company.
How to Get the Best Deals on Airline Tickets, According to Google Flights
3 hari lalu
This ticket booking recommendation is compiled after analyzing prices in 4,000 markets between 2021 and 2025 through Google Flights.
Today's Top 3 News: What to Know About Gemini AI's Viral Miniature Action Figure Generator
3 hari lalu
Here is the list of the top 3 news on Tempo English today.
A Transnational Investigation Reveals How Big Tech Lobby Operates in the Global South
4 hari lalu
Months of investigation involving more than 40 journalists exposes how technology giants deploy extensive lobbying strategies to shape regulations.
Nokia Opens New Research Campus in Finland for AI-Based 5G, 6G Technologies
4 hari lalu
The inauguration of Nokia's research and development center on September 5, 2025 will accommodate around 3,000 experts.
What to Know About Gemini AI's Viral Miniature Action Figure Generator
5 hari lalu
The craze began after Google rolled out Gemini 2.5 Flash Image (Nano Banana), a feature that turns photos and prompts into digital figurines.
Warner Bros Discovery Sues Midjourney for Superman Copyright Infringement
6 hari lalu
Warner Bros Discovery has filed a lawsuit against Midjourney for allegedly replicating popular characters such as Superman using AI.
UNICEF: Gaza is a City of Fear, Flight and Funerals
7 hari lalu
UNICEF reports that Gaza City, the last refuge for residents in the northern Gaza Strip, is rapidly becoming a place where children cannot survive.
