China’s extraction marketing campaign: A focusing on operation, not a curiosity
Anthropic’s disclosure that three China‑based mostly AI firms (DeepSeek, Moonshot AI, and MiniMax) ran greater than 16 million interactions by way of roughly 24,000 fraudulent accounts isn’t a narrative about mannequin misuse. It’s a story about focusing on. These campaigns went straight at Claude’s most delicate capabilities: agentic reasoning, device use, and coding. That isn’t random sampling; that’s structured assortment.
I’ve spent sufficient time on the earth of focusing on to acknowledge this sample instantly, and also you don’t want my degree of expertise to see it. When an adversary can observe a system at scale, they will map its strengths, seams, and predictable behaviors. China now has that behavioral telemetry for Claude, and they’ll use it to tune their very own methods and to form offensive operations in opposition to environments the place Claude‑like fashions are deployed.
And Claude isn’t the one system in China’s focusing on sights. The identical actors have used related excessive‑quantity extraction strategies in opposition to different frontier fashions, together with Google’s Gemini and OpenAI’s ChatGPT. They generate sufficient interplay knowledge to grasp how these methods suppose and the place they are often pressured.
