The Education edition
This is a real edition from 8 September 2026, at the Practical level, in plain language.
A court ruling on made-up sources, an AI pause in New York, and a deadline for your school's digital security.
Made-up sources in a paper count as fraud, even without intent
The highest administrative court says: it's about the outcome, not the student's intention.
The Administrative Jurisdiction Division of the Council of State confirms that made-up sources count as fraud. This applies even if the source says something different to what the student claims. It even applies when an AI tool made up the source and the student meant no harm. So an exam board doesn't need to prove it was done on purpose. What counts is the outcome, not the intention. A student can still appeal within six weeks to the Board of Appeal for Examinations, if the punishment seems too harsh.
This edition mainly gives you something to hold onto. The ruling on made-up sources means that with fraud, you only need to check if a source is real, not whether a student lied. The debate on AI at a young age, the drop in IT course enrolment, and Kennisnet's new self-assessment show that policy on AI in schools is very much still developing, from classroom use to digital security.
New York wants to keep AI out of the classroom for its youngest pupils, for a year
Public schools in New York City are considering pausing AI use by pupils up to secondary school age, for a year. That would affect almost 600,000 young pupils. Other American school districts are choosing to roll AI out fully instead. Read more →
OpenAI publishes its own research on ChatGPT and critical thinking
In late August, OpenAI set out what pupils and students gain, according to the company, when ChatGPT use is combined with training in critical thinking. The piece appears while schools worldwide are wrestling with their AI policy. Read more →
IT course enrolment down by more than twenty percent
Dutch universities have seen enrolment in IT courses drop by more than twenty percent since 2022. This comes from enquiries with umbrella organisation UNL. The likely cause: generative AI makes programming with good prompts more accessible. Read more →
Kennisnet renews its self-assessment for privacy and security
Kennisnet published a renewed tool on 3 September: the Zelfevaluatie Normenkader IBP (Self-assessment Privacy and Security Framework). It gives school boards step-by-step insight into their digital security, and lets them draw up an action plan straight away, with milestones and responsibilities. Read more →
Kennisnet is looking for schools to talk about the IBP website
Kennisnet is researching how the Normenkader IBP website can better match what schools need. You can sign up for a 60-minute online conversation, on 29 September or 1 October. You'll get 60 euros or a ticket for OnderwijsInzicht Focus IBP. Read more →
ChatGPT now falls under the strictest European rules for large services
The European Commission has designated ChatGPT, Reddit and Roblox as 'very large online services' under the Digital Services Act. They reach at least 45 million users a month in the EU. By January 2027 at the latest, they must meet stricter requirements. Read more →
Dictate your lesson prep in Gmail, Docs or Keep
Google is adding voice control to Gmail, Docs and Keep, built on Gemini 3.5 Live. You can dictate a document, search your inbox by asking a question, or turn loose thoughts into organised notes. Read more →
AI and tech, made simple.
This is a sample of the Education edition. Subscribers get it in their inbox every week.
Subscribe
inklaretaal · Amsterdam · © 2026
This is the generic Education edition. Subscribe and you get it made to measure: tuned to your level, your job and the topics that matter to you. That way every week gets a little more personal.
Subscribe →