Philosophers Debate the Role of AI in Education
amsterdam, donderdag, 17 juli 2025.
Two philosophers from the Philosophical Eleven discuss the advantages and disadvantages of artificial intelligence (AI) in education. Bas Haring, philosopher at Leiden University, sees AI as a useful tool that can support logical reasoning. Paul van Tongeren, emeritus professor in Nijmegen, is sceptical about the acclaimed importance of AI and argues that true knowledge only arises when it is understood and internalised by humans.
Philosophical Reflection on AI in Education
Bas Haring, philosopher and special professor of public understanding of science at Leiden University, views AI as a useful tool that can support logical reasoning. He suggests that AI bots can reason logically, suggest alternative perspectives, and check for gaps in reasoning. Haring compares AI to a ‘scattered old librarian’ who can guide students in their learning process [1]. Paul van Tongeren, emeritus professor of philosophical ethics in Nijmegen and Leuven, is sceptical about the acclaimed importance of AI. According to Van Tongeren, true knowledge only arises when it is understood and internalised by humans. He emphasises the difference between ‘external knowledge’ stored in data centres and genuine knowledge known by humans [1].
Advantages and Challenges of AI in Education
Haring reports a case of a student who quickly generated new ideas for her thesis through intensive communication with ChatGPT. This made him question his role as a thesis supervisor, but he stresses that caution is always necessary when applying new technologies in education. Both philosophers agree on the need to critically monitor the integration of AI in education and develop adapted forms of guidance and assessment [1]. Scientists from Radboud University and the University of Amsterdam warned in an open letter at the end of June 2025 about the use of AI in education, arguing that it could undermine critical thinking skills [1].
Historical Parallels and Serendipity
The debate on the role of AI in education is taking place within the framework of the Philosophical Eleven, an initiative of Trouw. Both philosophers draw a parallel with Socrates’ concerns about writing. Centuries before we felt such anxiety about AI and the internet, Socrates feared that the invention of writing would make us all stupid, as it would pour our memory onto paper. Van Tongeren points out that Socrates was wrong, but his argument holds true if the written texts remain unread. Writing has greatly expanded the possibilities for gaining insight, but only if the texts are read. Serendipity, the unexpected discovery of useful information during library searches, is another aspect discussed [1].
Practical Examples of AI in Education
In practice, various initiatives are already underway to integrate AI into education. The Nestas School Group (primary education) and the Portus School Group (secondary education) have received approval from NOLAI (National Education Lab AI) for the BEGR(AI)P project. This project will develop an AI-supported reading environment over three years, which will help students with comprehension and provide immediate feedback [2]. A student from a second-year MAVO class said: ‘I read it, but I don’t always understand what it says. I would like to learn how to understand texts more easily without just rereading them.’ [2]
AI Literacy as a Crucial Skill
AI is developing rapidly and finding its way into the classroom. This calls for fundamental adjustments in educational practices and the skills of teachers and students. AI literacy—the ability to engage with AI in a critical, responsible, and effective manner—is becoming increasingly important. A knowledge product from Npuls presents an evidence-informed framework that helps education professionals understand and promote AI literacy. The framework is based on an extensive umbrella review of existing scientific literature, supplemented with practical examples and feedback from education professionals through World Cafés [3].
Ethical Aspects and Privacy
The integration of AI into education also raises challenging questions about the role of algorithms and AI in the learning process. Technology can (unintentionally) clash with values. How do you handle the shifting balance between human and machine? And how do you ensure information security and privacy in your organisation? During the ‘Control Over Digitalisation’ course, organised by the PO Council, leaders, staff members, and school principals receive guidance to answer these questions. They learn how to safeguard public values in education and steer their digital ambitions [4].
International Dimension: Hidden Prompts in Scientific Publications
On an international level, there are also discussions about the impact of AI on scientific research. Some scientists have managed to embed instructions in AI software within their publications, aimed at generating positive reviews. These hidden prompts, often in white text on a white background or in a tiny font, enable AI tools like ChatGPT and Perplexity to generate positive reviews. Research by Nikkei has shown that seventeen articles with hidden prompts were found, primarily from researchers in Japan and South Korea. Tests by the HOP (Higher Education Policy Institute) have demonstrated that both ChatGPT and Perplexity respond to these hidden prompts, with ChatGPT initially signalling a ‘significant integrity violation’ but generating a very positive review after adjusting the prompt [5].