Security researchers have discovered a 'completely open' DeepSeek AI database that contained chat histories, APIs, and other ...

Chinese AI startup DeepSeek is facing a cyberattack that has disrupted services while its chatbot declines to discuss ...

Security researchers tested 50 well-known jailbreaks against DeepSeek’s popular new AI chatbot. It didn’t stop a single one.

2025年1月26日，拾象创始人兼CEO李广密，组织了一场关于 DeepSeek的闭门讨论会，嘉宾包括数十位顶尖AI研究员、投资人与一线 AI 从业者，围绕DeepSeek的技术细节、组织文化以及其出圈后的短中长期影响等，进行了探讨与学习。

2025年1月26日，拾象创始人兼CEO李广密，组织了一场关于DeepSeek的闭门讨论会，嘉宾包括数十位顶尖AI研究员、投资人与一线 AI 从业者，他们围绕DeepSeek的技术细节、组织文化以及其出圈后的短中长期影响等，进行了探讨与学习。

A security report shows that DeepSeek R1 can generate more harmful content than other AI models without any jailbreaks.

DeepSeek 以始料未及的速度引发了全球AI社区的狂热：R-1 的讨论还没降温，今天凌晨， DeepSeek 又发布了开源多模态模型 Janus-Pro。2 天前，拾象组织了一场聚焦于 DeepSeek的「Best Ideas 闭门讨论会」，在长达 3 个小时的讨论中，我们和顶尖AI 研究员、AI 开发者和投资人一起围绕 ...

China-based DeepSeek has exploded in popularity, drawing greater scrutiny. Case in point: Security researchers found more ...

The heightened drama around DeepSeek rests on a false premise: Large language models are the Holy Grail. On the contrary, the ...

Users are jailbreaking DeepSeek to discuss censored topics like Tiananmen Square, Taiwan, and the Cultural Revolution.

DeepSeek sparked hope for AI on cheaper chips with open-source code, but a security flaw now serves as a wake-up call for ...