OpenAI has begun implementing new safety measures across its ChatGPT platform aimed at preventing real-world violence. The company has placed new limits on weapons-related information for minors and introducing automated chat terminations for users showing warning signs.
Updates include restrictions on practical details regarding weapons and security systems for accounts estimated to be under 18 years of age, even when queries appear harmless.
OpenAI stated its age prediction system automatically moves users into the teen experience based on behavioral signals, while parents will soon receive notifications if a teen account is paused for repeatedly discussing violence.
For all users regardless of age, the ChatGPT-maker is restricting detailed reconstructions of past mass attacks, including information on weapons used and how they were obtained.
Its automated safety interventions can immediately end conversations when a user appears to be fixating on violence. Repeat violations can also lead to temporary account pauses.
OpenAI will begin publishing a regular Community Safety Report in early 2027, sharing data and findings on its safeguards and violence prevention efforts.
Source: Mobile World Live
Image Credit: OpenAI
Source: Tahawul Tech