China has temporarily disabled image-recognition features on major AI applications during its national college entrance exam period to prevent cheating. Popular AI tools from Alibaba, ByteDance, DeepSeek, Moonshot, and Tencent are blocking photo-based queries that could help students cheat on exam papers, demonstrating how governments can coordinate with tech companies to maintain academic integrity during high-stakes testing.
What you should know: China’s most popular AI applications have coordinated to prevent cheating during the gaokao exam period, which runs through June 10.
• Tools like Alibaba’s Qwen, ByteDance’s Doubao, DeepSeek, Moonshot’s Kimi, and Tencent’s Yuanbao are frozen during exam hours.
• The apps prevent photo-recognition features from working when they detect documents that look like exam papers.
• When users ask why features aren’t available, the tools respond that features are shut down to ensure fairness during exams.
The big picture: The gaokao exam, also known as the National College Entrance Exam (NCEE), is a multi-day process that millions of Chinese students undergo to earn college placement.
• This coordinated shutdown represents a systematic approach to preventing AI-assisted cheating that would be difficult to replicate in countries with less centralized control.
• The United States hasn’t implemented similar measures, partly due to less government control over tech platforms and the lack of centralized national exams beyond SATs or ACTs.
Meanwhile, AI is being used to catch cheaters: China is simultaneously deploying AI surveillance technology to monitor for traditional cheating behaviors during the same exams.
• According to China Daily, a state-run newspaper, the technology looks for suspicious behavior like whispering or repeatedly glancing at a neighbor’s paper.
• Footage flagged by AI will be reviewed after exam completion.
Contrasting approaches: The U.S. government is taking a different stance on AI in education, recently promoting integration rather than restriction.
• A White House executive order from April states: “It is the policy of the United States to promote AI literacy and proficiency among Americans by promoting the appropriate integration of AI into education, providing comprehensive AI training for educators, and fostering early exposure to AI concepts and technology to develop an AI-ready workforce and the next generation of American AI innovators.”
Why this matters: This represents one of the first large-scale coordinated efforts by AI companies to prevent academic cheating, highlighting how artificial intelligence tools are becoming sophisticated enough to require specific countermeasures during critical educational assessments.