High Court rules natural language opt-out of text and data mining insufficient
Background
The Danish Maritime and Commercial High Court issued a preliminary injunction on 22 October 2025 prohibiting ReData A/S from scraping data from BoligPortal's rental listing platform boligportal.dk. ReData scraped rental listing data directly from the publicly accessible platform boligportal.dk on a daily basis – the same underlying data that BoligPortal itself uses for its commercial Data Insights product.
The decision was appealed to the High Court.
The High Court's decision on TDM
A central question was whether BoligPortal had validly reserved its rights against text and data mining ("TDM") under Section 11b of the Danish Copyright Act, which implements Article 4 of Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on copyright and related rights in the Digital Single Market ("DSM").
BoligPortal had placed its reservation in its data and privacy policy, accessible via a footer link on the front page of boligportal.dk. The reservation was in HTML format, expressed in natural, human-readable language.
The High Court found that the most natural reading of the "machine-readable" requirement in Article 4(3) – read in light of Recital 18 of the DSM Directive – is that a reservation must not merely be registered by a machine, but must be interpreted and acted upon by it, so that automated scraping systems leave the content unused. This interpretation was supported by the Hamburg Court of Appeal's judgment of 10 December 2025 (LAION) and the Amsterdam District Court's judgment of 30 October 2024 (HowardsHome).
On this basis, and in light of the explanation by Boligportal's Chief Data Officer, the High Court concluded that a reservation expressed in natural, human-readable HTML in a data and privacy policy accessible via a footer link does not satisfy the machine-readable opt-out requirement under DSM Article 4(3).
The injunction was accordingly lifted.
Comments
The High Court's decision confirms that rights holders should implement machine-actionable mechanisms – such as opt-outs via robots.txt – that automated systems are programmed to recognise and act upon, if they wish to opt out of text and data mining. Opt-outs expressed in e.g. terms and conditions or "reservations" on a website in natural language do not constitute a sufficient opt-out.
The High Court's decision is the latest in a series of decisions from European courts of relevance for scraping, text and data mining as well as artificial intelligence - and many more are sure to follow.