Good morning. Monday energy — It’s Monday, July 20. Locked in and ready to go. Hugging Face just admitted its own security team had to sideline US frontier models mid breach and reach for a Chinese open source LLM instead, and that single decision says more about where AI power actually sits than any benchmark chart.
Today we are covering:
Hugging Face was breached by an autonomous AI agent, and its security team turned to a Chinese LLM after US models blocked their own investigation
AI powered “synthetic insider” attacks let North Korean operatives steal millions from over 100 US companies using deepfake identities
Ben Thompson argues the panic over Chinese open weight models is overblown everywhere except cybersecurity
