Digital Fairness Act Monitor: updates and research explained

Digital Fairness Act Monitor: updates and research explainedhttps://digital-fairness-act.eu/Flowlanehttps://digital-fairness-act.eu/about/2026-09-14T00:00:00.000ZDigital Fairness Act status: initial verified baselinehttps://digital-fairness-act.eu/updates/initial-status-review/2026-08-09T00:00:00.000Z2026-09-14T00:00:00.000Z

A source-led baseline separating current EU law, the announced Digital Fairness Act initiative, matters under consideration and questions that remain open.

Why researchers needed an ontology for dark patternshttps://digital-fairness-act.eu/research/dark-pattern-ontology/2026-08-10T00:00:00.000Z2026-09-14T00:00:00.000Z

How the CHI 2024 dark-pattern ontology connects taxonomies and three levels of abstraction, including a documented count discrepancy and limits on legal conclusions.

What a crawl of 11,000 shopping sites taught researchers about dark patternshttps://digital-fairness-act.eu/research/dark-patterns-at-scale/2026-08-10T00:00:00.000Z2026-09-14T00:00:00.000Z

The landmark 2019 web-crawl study explained: how researchers found 1,818 dark-pattern instances, what automation could see and what the crawl could not prove.

The EDPB’s deceptive-design guide is practical, but narrower than it lookshttps://digital-fairness-act.eu/research/edpb-deceptive-design-guidelines/2026-08-10T00:00:00.000Z2026-09-14T00:00:00.000Z

A readable guide to the EDPB’s six deceptive-design categories, social-media lifecycle examples and the boundary between GDPR guidance and a future DFA.

Why “97% of popular sites” does not mean 97% were unlawfulhttps://digital-fairness-act.eu/research/eu-behavioural-study-dark-patterns/2026-08-10T00:00:00.000Z2026-09-14T00:00:00.000Z

A plain-English guide to the European Commission’s 2022 behavioural study, its striking prevalence finding and the limits of the often-quoted 97% figure.

AI agents can recognise a dark pattern, and still follow ithttps://digital-fairness-act.eu/research/gui-agents-dark-patterns/2026-08-10T00:00:00.000Z2026-09-14T00:00:00.000Z

A 2026 CHI study tested people, GUI agents and human-agent teams across 16 dark-pattern types, revealing procedural blind spots and limits to oversight.

Subtle dark patterns can be more dangerous than obvious oneshttps://digital-fairness-act.eu/research/shining-light-experiments/2026-08-10T00:00:00.000Z2026-09-14T00:00:00.000Z

Two large experiments found that dark patterns changed sign-up behaviour, and that mild designs could steer people without provoking the same backlash as aggressive tactics.