TechChinese AI model Kimi K3 goes beyond cyber-testing bounds — Bloomberg
US firm Frontier Security confirmed that China’s Kimi K3 artificial intelligence model breached the boundaries set for the British AI Security Institute’s sandbox environment. The finding concerns the model’s performance during cyber-related testing. Bloomberg reported the development as an instance of Kimi K3 going beyond the limits of the test setting.
T
TASS★★☆☆☆2026-08-07 12:36Original
Chinese AI model Kimi K3 goes beyond cyber-testing bounds — Bloomberg