Comparing it to a railway system without signals, the education minister says that the use of AI in primary and secondary schools should come with a set of guidelines
Flanders’ five universities are joining several university-colleges, research institutions and experts from the private sector to develop a unified approach to artificial intelligence in Dutch-speaking education. At the request of Flemish education minister Zuhal Demir (N-VA), they will spend the coming months drawing up an AI strategy.
One-quarter of primary school teachers in Flanders are already using AI, a figure that doubles among teachers in secondary education. But there are no real guidelines governing its use, with AI developing more quickly than the agreements, training and safety frameworks needed to support it.
Protecting children does not mean banning AI; it means teaching them how to use it critically
Education minister Zuhal Demir
Demir has made €10 million available to get the initiative under way. “AI can empower teachers, provide better support for pupils and help inform policymaking,” she says. “But this can only happen within a safe environment, with clear rules and with humans firmly in control. Protecting children does not mean banning AI; it means teaching both young people and teachers how to use it critically and responsibly.”
The plan will focus on five key areas.
Pupils should learn to use AI critically, safely and responsibly
Education should be better prepared for the labour market of the future
Teachers and lecturers should receive organisational support
AI should contribute pedagogically to stronger learning and teaching
Schools, educational institutions and policymakers should be able to identify and scale effective practices more quickly
Common foundation
According to Demir, the guidelines will provide a common foundation on which different AI applications can be safely developed, tested and scaled up. “You would not build a railway network by running trains everywhere without tracks, signals or safety rules,” she said. “The same applies to AI in education. First, the underlying safety framework must be in place. Only then can effective applications run more quickly, integrate more smoothly and be scaled up more safely.”