Chinese AI Labs Harvested Claude, Leaked Own Secrets — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic's report reveals that major Chinese AI developers operated illicit proxy networks to harvest Claude's reasoning capabilities.
- The proxy operations unintentionally leaked sensitive military, government, and corporate secrets during the process.
Why it matters: Chinese AI developers attempted to reverse-engineer a frontier U.S. model's capabilities through covert proxy networks, but the operation backfired by exposing their own sensitive military, government, and corporate data — turning an intelligence-gathering scheme into an unintentional intelligence leak.
Ask SkimNews
