Event ·Multi-source

Moonshot AI 通过训练数据过滤削弱 Kimi K3.1 Pro 网络攻击能力

Moonshot AI 对 Kimi K3.1 Pro 采取了安全限制措施，方法是在训练阶段过滤数据，降低模型的网络攻击能力。此举的目的是防止模型被滥用来挖掘软件漏洞。报道来自越南媒体 VTV，聚焦中国 AI 公司在模型安全方面的操作。

Current read

Moonshot AI 给 Kimi K3.1 Pro 降低了网络攻击能力，做法是过滤训练数据，防止模型被拿去挖软件漏洞，了解下他们怎么做的。

11 sources34° AI heatLast updated 2026/10/25 12:27:00

Evidence

11 sources

Primary sourceVTV Công nghệ

2026/10/25 12:27:00View source

Related source 1Blognone

2026/10/24 04:29:56View source

Related source 2kimmonismus

2026/10/24 05:59:19View source

Related source 3Tech in Asia

2026/10/24 08:10:47View source

Related source 4Teortaxes

2026/10/23 16:32:20View source

Related source 5kimi_moonshot

2026/10/23 17:15:24View source

Related source 6berryxia

2026/10/23 18:20:36View source

Related source 7Decoder

2026/10/23 19:00:07View source

Related source 8rohanpaul_ai

2026/10/23 19:21:53View source

Related source 9marktechpost

2026/10/23 22:52:40View source

Related source 10Latent Space (swyx)

2026/10/24 07:18:29View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story