November 1, 2026
OpenAI announced the imposition of new security restrictions on the GBT Chat platform, targeting information related to weapons and violence-related conversations, as part of its efforts to enhance the safety of users, especially teenagers, and expand mechanisms for notifying parents.
According to CNBC, this step comes amid mounting global scrutiny about the risks posed by artificial intelligence-supported chatbots on young age groups, and the extent of the responsibility of developing companies to monitor potential threats and deal with them firmly.
The new measures include restricting the provision of practical information related to weapons and security systems that may be exploited for harmful purposes, provided that these restrictions are applied to teenagers automatically, regardless of the presence of indicators indicating violent intentions in them. Restrictions also include limiting the availability of details about previous mass attacks to all users, including the types of weapons used and methods of obtaining them.
In a related context, the platform will be able to automatically end conversations with various users when it detects any excessive and repeated preoccupation with topics of violence, while imposing a penalty of temporarily suspending accounts if this type of conversation is repeated.
OpenAI explained that it has developed a system to estimate the ages of users, to ensure that those who are likely to be under eighteen are transferred to the version intended for teenagers, with plans to officially inform parents if their children’s accounts are suspended as a result of repeated conversations related to violence. The actual implementation of these procedures is scheduled to begin in the coming weeks, and they will be subject to continuous development and improvement.
It is noteworthy that OpenAI revealed in its first report on teenagers’ use of the platform that their average daily use is less than 15 minutes. On the other hand, a report issued by the non-profit organization “Common Sense Media” indicated that there were gaps in the systems for protecting teenagers, indicating that the version designated for them was delaying in notifying parents, or failing to send alerts at all, when monitoring conversations related to suicide, self-harm, or eating disorders.
