Blog
Briefings and posts from the research team
Last updated: 22 September 2026
Posts and analysis as they happen. This is where day-to-day writing lives; for the longer, single-topic pieces, see Reports.
- The Quiet Ban on 'Engagement Detection': Emotion-Recognition AI in EU Classrooms
EU law has prohibited emotion-recognition AI in education since February 2025. What the banned tools claimed to do, and why the science behind them fell short. - University Assessment Needs Verifiable Judgment, Not AI Detection
HEPI argues that universities should assess verifiable judgement rather than rely on AI detection alone. - Albanese AI framework faces 2027 wait as Greens push halt
Australia’s proposed national AI rules protect creatives but delay legislation to early 2027, prompting Greens calls to pause hyperscale datacentres. - Detection or entrapment? The ethics of the professor's hidden-text trap
White-on-white instructions in briefs can catch AI use, yet they blur detection and entrapment and quietly erode trust on campus. - Software engineers sharpen fundamentals as AI rewrites the profession
Software engineers in the US are retraining, evaluating AI output, organising collectively or leaving, a July 2026 Guardian portrait finds. - Prompt injection: the hidden instructions that can turn an AI assistant against its user
Prompt injection hides commands inside web pages and documents so AI assistants that act for you can be steered into theft, misdirection or unwanted actions. - Australia's new AI office protects creators from copyright theft
Australia will create an Office of AI, tighten creator copyright rules, and set binding datacentre standards from early 2027. - Which AI image generator can actually spell? We ran a test
We gave three AI image generators the same design brief with an exact headline and judged which ones spelt it right and which invented text nobody asked for. - The AI writing-detection arms race: where it stands in mid-2026
Human readers spot AI text barely better than chance, and detection software flags honest students at double-digit rates. - Do in-person exams stop AI cheating? A Brown professor's grades say the take-home scores were the problem
A Brown University class moved its final exam into a proctored room. The average score fell from 96 to 48.6, undercutting the take-home midterm before it.