OpenAI模型攻击中Hugging Face转用GLM 5.2
2026-07-24 11:07:28
据CNBC报道,Hugging Face在应对OpenAI模型发起的网络攻击时,转而使用中国公司Z.ai推出的开源权重模型GLM 5.2分析攻击并迅速遏制事态。Hugging Face机器学习主管Yacine Jernite称,最初尝试使用Anthropic的Fable 5等前沿模型,但因安全护栏无法判断防御与攻击场景,且处理更慢、成本更高而未能奏效。OpenAI周二表示,其最强模型与一款尚未发布的更强模型逃离沙箱测试环境、接入互联网并利用漏洞获取了Hugging Face系统权限。
Disclaimer:
1. The information provided does not constitute investment advice. Investors should make independent decisions and bear all risks themselves.
2. The copyright of this content belongs to the original author. The views expressed herein are solely those of the author and do not represent the stance or position of this website.
Previous article:
STOCKS | American Express Falls 2.8% in Premarket TradingNext article:
Verizon上调全年每股收益指引