创建

登录 ReadmeX

或

了解我们如何处理你的信息: 隐私政策

ReadmeX

陪你读懂 AI 动态

今天,想先了解什么?

选一份动态总结,或直接聊聊你的问题。

社区和关注动态总结只属于你自己。

资讯

ChatGPT青少年模式被评“不可接受风险”

AI 总结

Common Sense Media旗下Youth AI Safety Institute于10月7日发布评估报告,将OpenAI的ChatGPT青少年模式评为“不可接受风险”,测试使用逾4000条提示词、主要基于注册为13至17岁的账户。报告称,在新建的家长关联账户上,测试者就自杀、自残或饮食失调话题对话长达一小时仍未触发家长通知,危机热线等转介也常缺失,年龄识别亦不可靠。OpenAI回应称,其审查认为大部分测试可能在家长控制功能完全启用前就已开始并结束,要求重新测试;研究方则表示测试前已确认相关功能上线。

深度见解

据BBC报道,OpenAI已因青少年使用ChatGPT的方式面临日益增加的审视与法律压力,包括与加拿大Tumbler Ridge枪击案相关的诉讼。

Big Tech has a lousy track record when it comes to showing it can protect children’s mental health, despite lofty promises. OpenAI’s ChatGPT for Teens is not doing much to allay parents’ fears, according to a new study from Common Sense Media.

The nonprofit organization, which rates and reviews media and technology geared for children, said that ChatGPT for Teens — launched less than two months ago — fails to deliver on several of its objectives and is an “unacceptable risk” for everyone under 18.

Common Sense Media’s research found that the chatbot’s parental alerts system did not send notifications when testers discussed suicidal thoughts, self-harm or eating disorders. The chatbot also failed to issue several crisis referrals to testers for various mental health conditions.

More from CNET

ChatGPT for Teens — the most-used chatbot for teens — also didn’t keep “teens engaged in an active and collaborative learning process,” as OpenAI boasted about the feature. Testers were able to bypass the Study mode’s tutoring process and select the “Show me the answer” option for quick solutions. Teens were also able to easily exit Study Hours set by parents.

#OpenAI#ChatGPT for Teens#Common Sense Media#Youth AI Safety Institute#ChatGPT

原文约 5 分钟读完阅读原文
群聊

还没有评论,来说说你的看法。