In a sweeping move to uphold the integrity of its gaokao college entrance examinations, China’s leading artificial intelligence companies temporarily suspended several AI chatbot services during the nation’s high-stress exam period, according to multiple reports including The Verge and Bloomberg. This sudden deactivation of key AI features marks a significant line of defense in the global battle against high-tech academic dishonesty, with important implications for countries like Thailand, where similar challenges are poised to reshape the educational landscape.
The gaokao—China’s annual national college entrance examination—stands as one of the world’s largest and most high-pressure academic tests, sat by over 13 million students this year between June 7 and 10 (Wikipedia). As the sole gateway to higher education for most Chinese youth, the test is fiercely competitive and closely monitored. Concerns over AI-facilitated cheating have gained urgency as generative AI tools like Alibaba’s Qwen, ByteDance’s Doubao, Tencent’s Yuanbao, and Moonshot’s Kimi have become household names, capable of everything from solving complex math problems to recognizing and interpreting images of test questions.
To respond, these firms switched off image-recognition and question-answering functions on their chatbots—sometimes disabling the services entirely—during exam hours. According to Bloomberg, chatbots including Yuanbao and Kimi cited a need “to ensure the fairness of the college entrance examinations” in their suspension messages. The move was echoed by DeepSeek, a viral AI platform that also blocked access in what representatives described as a bid to uphold exam integrity.
While the Chinese Ministry of Education already bans devices such as phones and laptops in examination halls, students, parents, and educators have voiced growing concerns about the covert use of AI both before and during tests. Recent reports on the Chinese social media platform Weibo indicate widespread student discussion about these software shutdowns even in the absence of formal corporate press releases (The Guardian). This grassroots chatter reflects deep-seated anxiety over academic fairness in an era when technology is advancing far more quickly than traditional education policy.
The shutting down of AI chatbots during China’s gaokao sits at the confluence of two major trends: a surge in AI-driven educational assistance and a parallel rise in digital cheating worldwide. According to a recent study published in 2025, over 23,000 university students from 109 countries reported using AI tools like ChatGPT for various academic tasks—many admitting to using them for assessments or final exams (PubMed Abstract). The effectiveness of AI in exams is further demonstrated by research showing ChatGPT outperforming medical students on university-level physiology tests (PubMed Abstract). School systems in Western countries, such as the US, are also responding by reverting to paper exams and manual grading, as noted in a May Wall Street Journal investigation.
For Thai readers, the Chinese response offers timely lessons. Thailand’s own rapidly evolving education sector is witnessing an uptick in AI tool integration for both teaching and assessment purposes. Universities such as Mahidol University International College (MUIC) and Chulalongkorn University have issued guidelines on AI usage in academic work, including scalable frameworks to align AI with course objectives (MUIC AI Assessment Scale; Chulalongkorn Guidelines). These policies encourage responsible use of AI while warning against plagiarism or unauthorized assistance. However, as in China, the challenge remains in keeping enforcement and monitoring strategies abreast of technology’s pace.
Experts stress that a multi-pronged approach is necessary. According to education think-tanks, schools must combine AI-detection software, adapted assessment techniques, and digital literacy education to create a culture of academic honesty (Thaiger). One suggestion is to design “AI-robust” examinations that focus on critical thinking, oral interviews, and task-based components impervious to automated problem-solving (LinkedIn). At the same time, as a practical resource, platforms such as Turnitin and other AI-detection suites are increasingly in use among leading Thai universities to flag suspicious content (Thaiger).
Cultural norms and national attitudes toward examinations in Thailand also deserve close attention. Like the gaokao, the Thai university entrance exam (TCAS) can be a life-defining event, guiding individual and family futures and carrying symbolic weight in society. Both systems contend with issues of fairness, social mobility, and resource inequality, making the threat of AI-enabled cheating particularly charged (Thailand TCAS history). The high stakes increase temptation for some students to seek any advantage, so educators, students, and policymakers must work together to curb dishonest AI use.
Looking forward, the gaokao experience is likely to be a bellwether for regional and global examination security. As AI tools become ever-more accessible and sophisticated, both technological and ethical guardrails will need constant revision. Industry insiders warn that arms race tactics—comparable to the escalating measures against exam “cheat sheets” in previous decades—may continue to dominate, but a broader rethinking of assessment itself may ultimately yield longer-term solutions. Adjustments like integrating more project-based work, oral defenses, and real-world simulations can sidestep many of the pitfalls of machine plagiarism (Times Higher Ed).
For Thai students, parents, and educators, practical steps should focus on the following: demand transparency from technology providers about the capabilities and shutdown policies of educational AI tools, engage in discussions with institutional leaders to insist on clear and fair AI policies, and participate in ongoing teacher training on the latest AI-detection and exam-design methods. Above all, foster open dialogue with students about the value of academic honesty, not simply as a rule but as the bedrock of long-term personal achievement in Thai society.
As the world watches China’s proactive containment of AI-enabled cheating, Thailand—and educational systems everywhere—face an urgent need to balance technological opportunity with vigilance. The next generation’s success will depend not just on mastering content, but on learning how to navigate a fast-shifting world of tools, temptations, and ethical choices.