Despite US export controls, Chinese AI labs are reportedly training models on Nvidia's Blackwell hardware, with its fingerprint detected in models from DeepSeek, Moonshot AI, MiniMax, and Ant Group. Meanwhile, Anthropic is implementing behavior standards to protect its Claude models from insults and rough treatment by users, as a significant portion of AI users engage with bots in a hostile manner. In other news, AMD is preparing to launch its Zen 6 cores, promising faster clock speeds and improved IPC gains. AI
IMPACT Chinese AI labs may be circumventing export controls, while Anthropic is proactively managing user interaction with its models.
RANK_REASON The article is a news digest covering multiple unrelated topics, including AI hardware usage, AI model behavior standards, and CPU development, rather than a single focused event.
- Acronis
- AMD
- Ant Group
- Anthropic
- Blackwell
- Claude
- DeepSeek
- EQT Partners
- Jeff Moss
- Jensen Huang
- Microsoft
- Microsoft SharePoint
- Mikko Hyppönen
- Moonshot AI
- Nvidia
- Zen 6
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →