OpenAI says ChatGPT can automatically terminate conversations fixated on violence, and promises to regularly publish community safety reports
OpenAI said in a blog post on October 30 that ChatGPT can automatically terminate the current conversation when the dialogue shows that a user is fixated on violence or has thoughts of seriously harming others, and the same account may be temporarily frozen after being interrupted for this reason multiple times. OpenAI has also tightened restrictions on reviewing details of past mass attack incidents and limited information related to guns and weapons that teenagers can access. OpenAI promised to regularly publish a "Community Safety Report," with the first issue to be released in early 2027, covering safety data for the second half of 2026, and updated every six months thereafter; in the third quarter of 2026, it had already referred more than 200 case leads to law enforcement.
Why read this
AI score 65
The original text lists specific mechanisms such as automatic termination of violent conversations, account freezing, and parental notification, which can be used to understand the direction in which the platform's safety policies are tightening.
Topics
Body · AI translation
OpenAI says ChatGPT can automatically terminate conversations that show a violent obsession, and promises to regularly publish community safety reports
2026-10-31 07:09 IT Home - Qincang (intern)
Thanks to IT Home user Huanan Wuyanzu for the tip!
IT Home, October 31 news: OpenAI published a blog post on October 30 local time stating that maintaining community safety has always been OpenAI's top priority. Its daily work includes proactively reporting imminent violent threats to law enforcement. In the third quarter of 2026 alone, OpenAI handed over more than 200 case leads to law enforcement agencies.
OpenAI promises to regularly publish a "Community Safety Report," systematically disclosing progress in violence prevention, other community safety topics, and the actual effectiveness of various protective mechanisms. The first report will be officially released in early 2027, covering safety data for the second half of 2026, and will be updated every six months thereafter.
IT Home has learned that OpenAI has added additional restrictions on how ChatGPT answers questions about past mass violent attacks and the perpetrators: even if the relevant information has already been publicly released online, the system will be strictly controlled and provide only extremely condensed information. At the same time, the public can still use ChatGPT to understand the full picture of such incidents, the impact on victims and communities, and learn how to identify warning signs of danger to prevent future violence.
But OpenAI has further tightened its safety policy, resolutely restricting detailed retellings of past mass attacks, including the specifications of the weapons involved, how they were obtained, and comparisons of the scale of harm that could fuel the perpetrators' notoriety. Regardless of the user's age or the motive behind the question, the above restrictions apply without distinction.
In addition, when a conversation shows that a user is beginning to obsess over violence or has thoughts of seriously harming others, ChatGPT can automatically terminate the current conversation, thereby immediately blocking this psychological and behavioral tendency. If the same account has its conversations forcibly interrupted multiple times for this reason, OpenAI may temporarily freeze the account.
OpenAI also announced that it has added stricter thresholds for information related to guns and weapons that minors can access; even seemingly harmless questions will be restricted. Even if the user shows no violent intent, the system will comprehensively restrict the provision of practical knowledge about weapons that could be misused to breed violence, as well as information on bypassing security systems.
OpenAI is currently stepping up development of a new feature to promptly notify parents when a minor's account is temporarily frozen due to repeated discussions of violent content.