Anthropic 国际负责人称开源权重模型风险更高:因无法监控私有部署与移除安全护栏
原标题:Bloomberg: Anthropic international chief Guillaume Princen said open-weight AI models carry more saf...
42AI 研判分
Anthropic 国际业务负责人 Guillaume Princen 在都柏林 AI 峰会上指出,开源权重模型本质上具有更高的安全风险。其核心论据在于控制权缺失:一旦权重公开,开发者无法阻止用户移除内置的安全护栏,也无法监控或召回在私有系统中运行的模型副本。
报道正文
Rohan Paul (@rohanpaul_ai) · 2026-10-14 11
Bloomberg: Anthropic international chief Guillaume Princen said open-weight AI models carry more safety risk by definition.
“I can’t talk for open models, but by definition, there’s more risk with them,” Princen said at the International AI Summit in Dublin on Wednesday.
His concern centres on control: once open weights are released, safeguards can be removed and copies can run on private systems that no developer can monitor or recall.
