Book

Description

Between 2026 and 2027, the obligations of the EU AI act take effect on a staggered calendar. Most of the organizations they bind do not yet know they are bound. This volume consolidates three pieces of independent research into a single argument about how regulation (EU) 2024/1689 actually allocates its weight. It opens with an empirical screening of fifty AI companies across four sectors and fifteen-plus European jurisdictions, scored against the regulation's six classification gates. Seventy-four percent trigger high-risk classification. Ninety-six percent hold no public regulatory position. None of the fourteen companies carrying existing ce marks or banking licences have mapped the AI act layer on top of them. From that provider-side gap, the analysis follows the cascade downstream. A small number of high-risk providers serve thousands of European deployer organizations who inherit article 26 obligations without recognizing it. Three segments are mapped paragraph by paragraph: human-resources, healthcare, and retail-banking deployers, each with its own path to discovery, article 27 fria logic, and article 25 reverse-bascule check. The closing study reframes the whole. Set against the ai-futures predictive frame of the 2010s, it argues that the deployer-side documentation regime created by the regulation is a third structural pole the extinction-or-immortality binary never contained: slower, regulated, and already in force. A constrained, falsifiable prediction closes the book. The thesis throughout is one LINE. Classification comes first, and everything else follows from it.