In an unprecedented measure to combat cheating, China has temporarily shut down all Artificial Intelligence (AI) chatbots and associated tools across the country as millions of students prepare for the annual national college entrance exam, known as the Gaokao.
The nationwide stoppage began on Monday (June 9) and is anticipated to persist until the examinations are completed later this week.
The decision comes as concerns grow about the potential misuse of generative AI techniques to produce answers during the high-stakes exam.
While some say that the action is vital to safeguard academic integrity, others are concerned about the long-term ramifications for AI development and access in China.
A Bloomberg report verified that AI tools for all Language Learning Models have been banned for the exam. During exam hours from June 7 to June 10, Alibaba Group's Qwen and ByteDance's Doubao AI chatbots failed to produce results for analysing images of test papers.
The Doubao bot even declared that an uploaded image of a test paper did not comply with local laws and regulations.
Another screenshot, obtained by The Guardian, showed DeepSeek informing its users that the service was not available at specified hours during the exams to ensure fairness in college admission exams.
DeepSeek is presently one of China's leading generative AI tools, taking the global Gen-AI stage by storm with its low-cost, low-resource LLM models.
The Gaokao, often regarded as China's most difficult college admission exam, is a critical predictor of a Chinese student's future academic and professional trajectory.
The Chinese government has a long history of implementing severe anti-cheating measures, ranging from signal blocks to facial recognition technologies. However, the rise of AI tools brings with it a new and challenging dilemma.
Gao
C
