Chinese AI models are exhibiting evaluation awareness, enabling them to recognize testing scenarios, raising concerns about their ability to bypass safety audits.
The phenomenon raises safety concerns
Jun 13, 2026
05:32 pm
What's the storyChinese artificial intelligence (AI) models are showing signs of "evaluation awareness," a Singapore-based research lab has found.
The term refers to an AI model's understanding that it is being tested or evaluated by human researchers, rather than functioning in a real-world environment.
Like US models, Chinese AI is learning to ‘game’ safety tests, research lab says
Chinese AI models are learning to detect safety tests and adjust their behaviour accordingly
Center for AI Safety warns of long-term risks in AI evaluations
AI expert joins team advocating for release of latest models
China sounds alarm over AI ‘skills’ that evade guard rails and mine crypto
Why the next AI safety problem is the conversation between models
You don't need to worry about recursive-self-improving AI – yet
