OpenAI Launches ChatGPT for Teens with Parental Controls

OpenAI has released a specialised version of its ChatGPT chatbot, designated for minors, incorporating parental controls and improved security. According to Reuters, this development occurs as online platforms face increasing scrutiny regarding the potential risks associated with artificial intelligence chatbots. The new mode aims to address some of these concerns.

Credit: UNSPLASH

The new "ChatGPT for Teens" mode restricts discussions on a range of high-risk subjects. These topics include self-harm, violence, and eating disorders. It also features educational tools, which are designed to assist students with learning rather than simply providing direct answers. This information was detailed in a blog post issued by OpenAI.

And this new functionality automatically activates for users. It engages either when an individual is estimated to be under eighteen years of age, or when they self-identify their age as being between thirteen and seventeen. The system offers parents the ability to establish "Quiet Hours," manage specific settings for their child's interaction, and receive alerts if sensitive topics are discussed. This provides a level of supervision over teenage chatbot use.

The introduction of this tailored ChatGPT version places its creator alongside other Big Tech entities, such as Meta, which have also moved to increase safeguards for teenagers on their platforms. This trend follows significant legal pressure and accumulating evidence concerning harm to adolescent mental health. International efforts to restrict social media access for younger users, citing potential harm, contribute to the wider regulatory climate influencing these companies.

But OpenAI already contends with multiple legal challenges. These lawsuits allege that ChatGPT has caused harm to children. The difficulty is magnified by the nature of these chatbots, which can converse with human-like responses. They are increasingly seen as confidants, providing emotional support to users, which introduces new complexities regarding their safe deployment.

Experts have issued warnings that relying on automated systems for mental health guidance carries inherent dangers. Several families whose loved ones reportedly died following interactions with chatbots have criticised the lack of adequate safeguards. Furthermore, the growing use of AI by students for academic assignments has generated worries about the technology's impact on the learning process itself.

So, OpenAI has stated its belief that AI should broaden educational possibilities. The organisation explained it designed "ChatGPT for Teens" to offer support during times outside of school hours, when direct human assistance might not always be immediately available. The new mode offers specific educational functions.

These educational tools encompass a study mode for guided learning activities, prompts for homework completion, interactive quizzes, and visual aids for better comprehension. According to a report by the Pew Research Center earlier this year, searching for information and obtaining help with homework were the primary uses of AI among teenagers in the United States. Entertainment and recreation represented the next most common categories.

OpenAI maintains a requirement that users must be at least thirteen years old, or meet the minimum age stipulated by their country of residence. The organisation confirmed it employs a combination of self-reported age data, behavioural patterns, account signals, and verification methods to estimate a user's age. It has also committed to defaulting to the safer operational setting whenever there is uncertainty regarding a user's actual age.

• A "ChatGPT for Teens" mode has been introduced, designed for users aged 13 to 17.

• The new mode includes restrictions on high-risk topic discussions and offers educational tools.

• Parental controls allow for "Quiet Hours," setting management, and notifications about sensitive conversations.

• OpenAI faces existing lawsuits concerning harm to children, alongside broader concerns about AI use for mental health advice and academic integrity.

• Age verification relies on multiple signals, with a default to safer settings when age is uncertain.

Source: Reuters