The EU AI Act's transparency provisions took effect August 2026, requiring disclosure for AI-generated public-interest text. The law carves out an exemption for content that has undergone human editorial review and for which a named party holds accountability. Nielsen Norman Group is publishing this because they qualify for that exemption, and they want you to know exactly why.
NN/G uses AI for clarity, formatting, and critique during editing. AI may contribute words and ideas. But every article since 2013 has passed through human editorial judgment: topic selection, field relevance, unique perspective, and final accountability. The editor states it plainly: people with domain expertise, not the model, stand behind what gets published.
The piece is worth reading in full because it details the actual editorial workflow, not just the policy position. The exemption logic matters: the EU is not banning AI assistance, it is requiring a named human to be on the hook. NN/G is using that legal framework to explain what editorial responsibility actually looks like in practice, step by step.
[READ ORIGINAL →]