Experts urge caution in using AI tools for university applications

Amid widespread anxiety related to complex admission rules, many parents and candidates preparing college applications after the national college entrance exam, or gaokao, are turning to artificial intelligence tools in the hope that algorithms can instantly generate the "optimal plan".

However, education experts and teachers are warning that blind reliance on such AI services may backfire.

Li, whose son took this year's gaokao, told China National Radio that she purchased a digital card for an AI-powered consulting service from an online platform for 300 yuan ($44), far cheaper than traditional consulting services.

Another parent, surnamed Pan, purchased cards from several providers and found that each online platform generated vastly distinct recommendations.

"The efficiency of AI screening is indeed high, but the divergent results make me uneasy about actually submitting the applications," she said.

The Ministry of Education has repeatedly warned families to stay vigilant against various fraud schemes during the application and admission season. In an alert issued during the 2025 college admission period, the ministry cautioned that some individuals and organizations spread false information, heighten anxiety and lure parents into purchasing expensive consulting services.

The ministry stressed that many so-called experts are temporary hires with no real credentials, noting that no professional certification of this kind has ever been issued by relevant authorities.

Xu Qicheng, a senior teacher at a high school in Tongling, Anhui province, who has guided college applications for years, said the data used by AI systems often lags behind current developments.

For instance, some universities have expanded or reduced enrollment for certain majors this year, and the popularity of majors continues to shift rapidly. Xu added that newly established programs introduce further variables that AI systems cannot fully account for.

"Many people are using the same AI software. Students with identical scores receive basically the same recommendations," Xu said.

Li Qing, director of the admissions office at Beijing University of Chemical Technology, stressed the importance of verifying data sources.

"You must add a reminder to obtain relevant scores from official websites. If you don't, the AI might scrape random data," he said.

Chen Hao, an associate professor at the School of Artificial Intelligence at Xidian University, said AI is not meant to replace human decision-making.

Its strength lies in information retrieval, organization and comparison, he told Xi'an-based Sanqin Daily.

"In a sense, the rationality of AI-generated plans depends on the breadth, accuracy and authority of the data it collects," Chen said.

He suggested that candidates adopt a "human-machine collaborative" approach, using AI as a reference while retaining final control over decisions.

Chen advised students and parents to carefully verify every recommendation by cross-checking official sources, including universities' historical admission scores, rankings and program information. Doing so can help reduce the impact of AI "hallucinations" and data inaccuracies.

Yang Weian, deputy dean of the faculty of education at Northeast Normal University, called on families to apply caution when considering AI marketing claims.

AI should be used to narrow the pool of potential institutions by inputting objective criteria such as scores, provincial rankings and subject requirements, allowing algorithms to quickly eliminate clearly ineligible options, he said in an interview with Guangming Daily.

At the same time, students should further refine the AI-generated list by incorporating personal preferences, including preferred cities, academic interests and career aspirations, Yang said.