OpenAI strengthens ChatGPT safety for teenagers

What has OpenAI announced?

OpenAI has introduced a comprehensive set of measures to make ChatGPT safer for teenagers, as detailed in its official blog on [date]. The main new features include age-based protections, parental controls, educational tools, and partnerships with experts. Specifically, the model now detects and blocks inappropriate content for minors, adapting its responses to age ranges (13-15 and 16-17). Parents can monitor usage through a dashboard that shows conversation history and allows setting time limits. Additionally, features like step-by-step explanations and summaries of complex concepts are incorporated, designed to support learning. OpenAI has worked with organizations such as Common Sense Media and child psychologists to define the safeguards, according to the blog. The rollout will be gradual over the coming months for users aged 13 to 17, though the company has not specified exact dates.

Why is this important?

Teenagers are one of the fastest-adopting groups of generative AI: a Pew Research Center study (2023) indicates that 40% of young people aged 13 to 17 in the U.S. have used ChatGPT, but they are also the most vulnerable to risks such as misinformation, inappropriate content, or technological dependency. This move by OpenAI sets a precedent in the industry, which until now had prioritized user expansion without specific adjustments for minors. The measure responds to growing regulatory pressure, especially in Europe with the Digital Services Act and in the U.S. with proposals like the Kids Online Safety Act (KOSA). It also coincides with research warning about gender and racial biases in AI models, which could disproportionately affect teenagers. OpenAI has stated that content filters are based on guidelines from the World Health Organization and the American Academy of Pediatrics, though technical details have not been published.

Consequences for the market and users

With this update, OpenAI seeks to differentiate itself from competitors like Google Bard (now Gemini) or Anthropic Claude, which still do not offer native parental controls. Bard allows parents to manage minors' Google accounts but without a specific AI dashboard. Claude, for its part, has focused on general safety without age segmentation. For schools and families, the tool becomes more attractive as a supervised educational resource: according to an EdWeek survey (2024), 60% of teachers would be willing to use ChatGPT if it had parental controls. However, doubts persist about the real effectiveness of filters: in independent tests, similar content moderation systems have failed up to 20% of the time (AI Now Institute, 2023). There are also concerns about the privacy of minors' data, as OpenAI will store conversations to improve the model, though it assures it will not use them for training without explicit consent. The U.S. Federal Trade Commission (FTC) has warned that companies must comply with the Children's Online Privacy Protection Act (COPPA), which requires verifiable parental consent to collect data from children under 13. OpenAI claims its service is designed for users over 13, but age verification remains a technical challenge.

What readers should know

The new features will be rolled out gradually over the coming months for users aged 13 to 17. OpenAI recommends parents activate parental controls from first use and review the responsible use guides the company will publish soon. It is important to note that no protection is infallible: a Stanford University report (2024) found that language models can bypass content filters through 'jailbreaking' techniques. Therefore, human supervision remains key. Parents should maintain open conversations with their children about the risks and benefits of AI. Additionally, OpenAI has announced it will collaborate with schools to integrate ChatGPT into the classroom under teacher supervision. On the regulatory front, the European Union is evaluating whether these measures comply with the requirements of the AI Act, which classifies AI systems for minors as 'limited risk' but demands transparency. Finally, market analysts predict this move could increase OpenAI's young user base by 30% in the next year, according to Gartner estimates.

"We believe teenagers deserve access to safe and educational AI, with appropriate protections for their age," states the OpenAI blog.