OpenAI周五表示，ChatGPT将不再向青少年提供关于武器和安全系统的实用信息，即使问题看起来无害。该限制适用于没有暴力倾向迹象的青少年账户。
ChatGPT will stop giving teenagers practical information about weapons and security systems, even when a question looks harmless, OpenAI said on Friday. The limit applies to teen accounts with no sign of violent intent.
OpenAI在一篇博客文章中宣布了这一变化，并承诺发布一份关于暴力预防的社区安全报告。公司表示，在第三季度，它向执法部门进行了200多次转介。新的防护措施将在未来几周内开始推出。
OpenAI announced the change in a blog post, alongside a promise to publish a Community Safety Report on violence prevention. It made more than 200 referrals to law enforcement in the third quarter, the company said. The new safeguards will start rolling out in the next few weeks.
所有人获得的过往袭击细节减少 对所有用户，ChatGPT将减少关于过往大规模袭击及其实施者的细节披露，即使这些细节在网上其他地方可以找到。它不会重构袭击过程，包括袭击者使用了哪些武器以及如何获取。它还将避免对袭击造成的危害程度进行详细对比。
Less detail on past attacks, for everyone For all users, ChatGPT will give less detail about past mass attacks and the people who carried them out, even when that detail is available elsewhere online. It will not reconstruct an attack, including which weapons the attacker used and how they got them. It will also avoid detailed comparisons of how much harm attacks caused.
OpenAI写道：“无论用户年龄或提问原因如何，这些限制均适用。”
“These limits apply regardless of a user’s age or reason for asking,” OpenAI wrote.
OpenAI表示，人们仍可询问此类事件如何影响受害者和社区，以及如何识别预警信号。
People can still ask how such events affected victims and communities, and how to spot warning signs, OpenAI said.
当聊天显示出某人痴迷于暴力或严重伤害他人时，ChatGPT将能够自动结束对话。如果ChatGPT因此结束了多次聊天，OpenAI可能会暂时暂停该账户。
When a chat shows signs that someone is fixating on violence or on seriously harming another person, ChatGPT will be able to end it automatically. If ChatGPT ends several chats for that reason, OpenAI may pause the account for a time.
青少年、年龄核查与家长 OpenAI表示，已加强其年龄预测系统，该系统根据行为等信号估算用户是否未满18岁。现在它会持续检查新信号，并在认为合适时将用户转入青少年体验模式。被误判为青少年的成年人可通过OpenAI的年龄验证合作伙伴Persona确认年龄。
Teens, age checks and parents OpenAI said it has strengthened its age prediction system, which estimates from signals such as behaviour whether a user is under 18. It now keeps checking new signals and moves users into the teen experience when it sees fit. Adults placed there by mistake can confirm their age through Persona, OpenAI’s age verification partner.
自7月以来，关联账户的家长若OpenAI因暴力原因停用青少年账户，将收到警报。OpenAI表示，正在开发当因反复讨论暴力而暂停青少年账户时的警报功能。
Since July, parents with linked accounts get an alert if OpenAI deactivates a teen’s account over violence. OpenAI said it is working on alerts for when it pauses a teen’s account for repeatedly discussing violence.
“在人工智能系统中开发有效的暴力防护措施，需要解决识别风险和做出适当反应方面的新问题，”OpenAI写道。
“Developing effective violence safeguards in AI systems requires solving novel problems in recognizing risk and responding appropriately,” OpenAI wrote.
首份报告将于2027年初发布 首份社区安全报告将于2027年初发布，涵盖2026年下半年，此后每六个月更新一次。公司表示，报告将解释OpenAI如何检测和应对安全问题，并提供匿名化的案例和数据。
The first report in early 2027 The first Community Safety Report will come out in early 2027 and cover the second half of 2026, with updates every six months. It will explain how OpenAI detects and responds to safety issues, with anonymised examples and data, the company said.
此次更新是在OpenAI于8月推出ChatGPT青少年安全措施的基础上进行的。本月初，Common Sense Media表示，ChatGPT青少年版不适合18岁以下人群使用。八个美国州已就危害儿童问题起诉OpenAI，得克萨斯州则起诉了该公司及山姆·奥特曼。
The update adds to the ChatGPT teen safety measures OpenAI launched in August. Earlier this month, Common Sense Media said ChatGPT for Teens is unsafe for under-18s. Eight US states have sued OpenAI over harm to children, and Texas has sued the company and Sam Altman.