Six months ago, I asked the simple question “3 years of ChatGPT – so what?”. I was reflecting the anniversary of ChatGPT on 30th of November. But it seems that the AI progress and impact has dramatically developed since then.
As you know, I’m a passionate promoter of Open Education as a broad philosophy for learning innovations. And I strongly believe in the power and impact of quality education. Therefore, I’m always reflecting how education and society are affected and challenged by Artificial Intelligence (AI).
Let me share my hopes and concerns with you and tell me your opinion on them:
Today, there is more an AI hype with promises for everything and all sectors while AI regulation is considered as hindering innovation and unique opportunities. But I believe that just the contrary is true: AI regulation is helpful and beneficial because it facilitates the responsible, safe and fair AI use and sets guardrails for all parties and stakeholders.
In the special sector education, we urgently need more AI regulation to safeguard our human rights, democracy, and societal learning objectives and equality! You can read my full argumentation in my new publication here: https://doi.org/10.1007/s43681-026-01092-5
However, governments worldwide are reducing AI regulations to reduce barriers for AI providers and to allow their expansion without strong and clear limits. As a consequence, we are starting experimentations with children without any proven evidence from scientific research that AI use in education is causing positive effects.
This spring, it happened that a leading AI provider (Anthropic) warned for their own new versions of Claude (Mythos 5 and Fable 5) due to their functionalities to find and exploit software bugs. They decided to delay their publication allowing software developers to solve identified vulnerabilities first what was appreciated broadly.
However, the American government has forced Anthropic now to shut down the access to Mythos 5 and Fable 5 without any precise reasoning. That is increasing the bad situation that single institutions decide what is happening with the AI deployment worldwide: First, the commercial AI providers and their single leaders were deciding and now, nations and their single leaders are following them.
That is challenging human rights and democracy in general what the Council of Europe also highlights in the new Handbook on AI and Human Rights (for which I was honoured to deliver the draft for the Section 4.7 “Education” together with Wayne Holmes). Not commercial interests of a few but our societal needs and benefits for the commons and all should guide our decision making on AI and education.
AI can and do harm everybody and in particular minorities and vulnerable groups. Children are especially exposed because they cannot understand the mechanisms and often do not critically reflect the AI results (and also most adults do not critically reflect AI outputs). The leak of Claude’s system prompts (more than 1000 lines of code regulating the AI results) and the daily changes of Grok (to deliver more or less faschistoid or sexistic outputs) caused by Elon Musk demonstrate that the AI providers are strongly influencing what AI systems are delivering as their results.
Thats means, only a few people (almost only white men) are deciding how AI systems are answering human questions as well as requests by other AI systems. Therefore, if we want to use AI in education, we need public AI systems and infrastructure independent from commercial interests. In addition, we need education about AI for teachers, for students and for the whole population to facilitate critical reflections and building of AI competences and AI literacy. But both requires a lot of time and huge ressources.
Therefore, we should broadly discuss and implement regulation of current AI use in education that is only experimentation in most cases at the moment. We need to reflect why we are using AI, what are our learning objectives for education with AI and whether we need to use AI in a specific situation at all. For the scientific research and analysis, we have established the Critical Studies on AI and Education (CSAI-ED) last Friday and you are most welcome to join!
And for the broad and global discussions, we need to join forces: A movement worldwide is required NOW to launch critical debates among all stakeholders and to achieve changes and regulations of our currently unregulated experimentations. AI should not serve commercial interests and manipulations but positively contribute to our society and our common good. If you want to support me, drop me a line and have a look on: https://act-NOW.eu
Again, I’m very curious about your personal opinions and changes:
What do you think about it and what happened to you during the last six months?
Thank you for sharing your thoughts and let us keep in contact!
I’m looking forward to a better future through Open Quality Education including AI use in innovative learning processes for All and to our close collaboration for realising it!
If interested, read all my latest news here: https://opening-up.education/news
Finally, I wish you an amazing and relaxing holiday season!
I’m going to attend the 27th International AIED Conference in Seoul now (https://aied-conference.org/2026) with our 4th Ethical AI&ED workshop (https://aied2026.learning-innovations.eu/) and co-chairing the PIP Track (https://aied-conference.org/2026/call-for-paper/pip): Drop me a line if we can meet there!