Week of August 17, 2026
🔴 Must Know
The EU AI Act split in two — disclosure rules are live, the hard part moved to 2027.
On 2 August, Article 50 transparency obligations took effect across the EU: any chatbot, agent or avatar interacting with students must clearly disclose that it is not human, and AI-generated or manipulated content must be labelled. Separately, the “Digital Omnibus on AI” (Regulation (EU) 2026/1744, in force 27 July) pushed the Annex III high-risk obligations — covering AI used in admissions, grading, exam proctoring and behaviour monitoring — from August 2026 back to 2 December 2027. Penalties for transparency breaches can reach €15m or 3% of global turnover.
Why it matters: if you enrol EU-resident students or use vendors serving EU markets, the disclosure duty applies now regardless of where your campus sits. The sixteen-month reprieve on admissions and assessment AI is real breathing room — but it is a deferral, not a repeal, and it is the right window to get a vendor inventory done unhurriedly.
New South Wales moves to review — and possibly restrict — take-home assessment.
On 10 August, Deputy Premier and Education Minister Prue Car directed the NSW Education Standards Authority to urgently review AI’s impact on the Higher School Certificate, including a possible moratorium on unsupervised take-home assessments, and to develop a common approach to detecting inappropriate AI use.
Why it matters: this is the first time a major curriculum authority has put a whole assessment format on the table rather than tightening rules around it. Any school running coursework, extended essays or portfolio-weighted assessment — IB, A-level, AP alike — should expect this argument to arrive in its own jurisdiction and should know now what proportion of its grades depend on unsupervised work.
Subscribe to continue reading
Subscribe to get access to the rest of this post and other subscriber-only content.