OpenAI to add parental controls to ChatGPT

OPENAI uses the parents concerned directly, because the AI ​​giant announces plans for a new series of parental surveillance features.

The company explained in a new blog article that it was going ahead with more robust tools for parents who hope to limit unhealthy interactions with its chatbot, because Openai faces its first unjustified death trial after the suicide death of a Californian adolescent.

The features – which will be published with other mental health initiatives in the next 120 days – include the account link between parent and adolescent users and a closer socket on chatbot interactions. Caregivers will be able to define how Chatgpt reacts (in accordance with the “age” parameter of the model) and deactivate the history and memory of the cat.

OPENAI also plans to add parental notifications that Flag when Chatgpt detects “a moment of acute distress”, explains the company. The functionality is still in development with the panel of experts from Openai.

In addition to the new options for parents, Openai said that it would expand its network of world doctors and its real -time router, a feature that can instantly change user interaction to a new cat or a new model of reasoning according to the conversational context. OPENAI explains that “sensitive conversations” will now be transferred to one of the company’s reasoning models, such as GPT-5-Thinking, to “provide more useful and beneficial answers, regardless of the model of a person selected for the first time”.

In the past year, AI companies have been examined more in -depth for not having responded to security problems with their chatbots, which are increasingly used as emotional companions by young users. Safety railings have been easily proven to be jailbreak, including the limits of how the chatbot responds to dangerous or illicit user requests.

Parental controls have become a first step by default for technological and social companies that have been accused of exacerbating the adolescent mental health crisis, allowing children of sexual abuse and not to address online predatory actors. But these features have their limits, according to experts, based on the proactivity and energy of parents rather than that of companies. Other children’s safety alternatives, including applications and online verification restrictions, have remained controversial.

While the debate and concerns are evolving concerning their effectiveness, AI companies have continued to deploy additional security railings. Anthropic recently announced that his chatbot Claude would now automatically end up harmful and abusive interactions, including sexual content involving minors – while the current cat becomes archived, users can still start another conversation. Faced with growing criticism, Meta has announced that it limits its IA avatars for teenage users, a provisional plan that involved reducing the number of available chatbots and forming them not to discuss subjects such as self -control, disorderly diet and inappropriate romantic interactions.