Anthropic 警告開源 AI 模型存在網路攻擊風險
anthropic 稱中國公司 z.ai 的開源模型存在重大網路攻擊風險。這加劇了關於開源前沿 AI 模型的爭論。
anthropic recently released a report arguing that a powerful new model by z.ai poses a major cyberattack risk due to its free availability for download and modification. this has reignited the debate over open frontier ai models, with critics warning of potential misuse by bad actors and supporters arguing that they prevent a handful of companies from controlling the technology. the report is the latest development in the ongoing discussion around the benefits and risks of open ai models.