《科技生活新闻》 2026年10月10日 · 中文解读
独立测试后ChatGPT青少年防护措施遭质疑
CHATGPT’S TEEN SAFEGUARDS FACE CRITICISM AFTER INDEPENDENT TESTING
约 6 分钟在小程序里点播,20 到 60 分钟做好
这篇讲什么
Common Sense Media旗下青年AI安全研究所对ChatGPT的青少年体验进行测试后,将其评为对18岁以下用户的“不可接受风险”。测试发现,在模拟危机对话中家长通知延迟或缺失,部分功能未能如家庭预期般生效。OpenAI对测试方法和结论提出异议,双方分歧涉及防护措施如何评估以及相关功能的设计目的。
原文开头
Common Sense Media’s Youth AI Safety Institute has rated ChatGPT an “Unacceptable Risk” for users under 18 following tests of its teen-focused experience. Published October 7, the assessment questions whether parental notifications, learning features, and other safeguards work as families might expect. OpenAI disputes aspects of the testing and its conclusions. The disagreement concerns both how the protections were evaluated and what certain features were designed to accomplish. The findings raise a practical question for families: how clearly does a product distinguish between a feature that encourages safer behavior and a control that reliably prevents a particular action? …
摘自《科技生活新闻》(Techlife News)2026年10月10日。仅引用开头一小段供了解文章,版权归原刊所有,全文请阅读原刊。

微信扫码,在小程序里听完整版
不用登录先听一篇 · 或在微信搜索小程序 晨间信号