A think tank on Tuesday urged the government to formulate frameworks for using artificial intelligence in schools, after a survey found most teachers and students already rely on AI tools daily.

The Our Hong Kong Foundation (OHKF) polled 1,200 teaching staff and students, mostly from secondary schools, on their use of AI in teaching and learning from July to December.

More than 90 percent of teachers said they used AI tools, mainly to assist with teaching and handle administrative work.

AI adoption was even higher among students, as 95 percent said they used it for tasks such as searching information and completing homework.

The foundation said teachers are generally worried that over-reliance on AI could weaken students’ development, especially in problem-solving and critical-thinking skills.

Both students and teachers also expressed concern about privacy and data security when using AI tools.

Teachers lack the ability and knowledge to properly guide students in using AI, the think tank said, pointing to survey results showing teachers rated their own AI proficiency even lower than the ratings students gave themselves.

The foundation said the government should formulate two key frameworks for AI use in education: one to guide teachers on instructing students in proper, ethical AI use, and the other to introduce a progressive AI literacy curriculum for students to build understanding step by step.

Deputy research director Victor Kwok said Hong Kong can take reference from frameworks in other places, such as those formulated by Singapore or the Organisation for Economic Co-operation and Development (OECD).

“It's practically impossible to have very binding frameworks,” he said.

“But just like we outlined the case for OECD they would throw out some scenarios of where we can draw parallels in terms of, using the same idea of how to use those tools and then the teachers can follow.”

The think tank also said the Education Bureau should establish a one-stop AI education resource platform to reduce repetitive resources scattered among schools, which causes wastage.

On-the-job AI training should cover all subject teachers instead of just targeting science teachers, while trainee teachers should also receive training in AI literacy and pedagogical applications, it added.