# Nathan Lambert：闭源模型才是AI风险冰山主体

- 来源：Nathan Lambert (@natolambert)
- 发布时间：2026-09-18 20:02
- AIHOT 分数：45
- AIHOT 链接：https://aihot.news/items/cmu6xo66a0iu7rowk7ieljyv3
- 原文链接：https://x.com/natolambert/status/2100918359604724059

## AI 摘要

随着 Claude 对 OpenAI 的外部攻击事件，闭源模型依然是 AI 风险冰山的主体，而非开源模型。它们 1) 更容易上手，2) 能力更强，3) 发布时防护措施存在漏洞。而将开源模型微调用于特定攻击则更难。

## 正文

With the external hack of OpenAI via Claude, closed models continue to be the tip of the iceberg on AI risks, not open models. They have been 1) easier to get started with, 2) more capable & 3) shipped w/ leaky safeguards. Finetuning open models to specific attacks is harder.