This browser is not actively supported anymore. For the best passle experience, we strongly recommend you upgrade your browser.
| 2 minutes read

AI & the UK’s proposed text and data mining exception - some observations

In its recently published consultation response, the UK Intellectual Property Office signalled its intention to introduce an expansive copyright and database right exception to facilitate text and data mining for any purpose (so including commercial text and data mining). Notably, the text and data mining activity cannot be subject to any licence fee, there is no indication that right holders will be entitled to equitable remuneration, and it will not be possible to contract out of the exception.

The announcement of this intended exception coincides with the recent publication by the UK Government of its policy paper on AI Regulation in which the Secretary of State for DCMS, Nadine Dorries, stated “I want the UK to be the best place in the world to found and grow an AI business and to strengthen the UK’s position so we translate AI’s tremendous potential into growth and societal benefits across the UK”. If this ambition is to be met, then AI systems will require input data to evolve and deliver these anticipated societal benefits and a permissive text and data mining exception may help minimise some of the legal risk associated with the use of input data in AI training.

Ostensibly, the proposed exception is unlikely to have a significant impact on rightholders who operate subscription or ‘pay for access’ models as the Intellectual Property Office has made it clear that a user is required to have lawful access to the text/data they propose to mine and interrogate. It is less clear what the effect will be on, for example, freely accessible websites or databases and the terms of use associated with those – for example, if a particular term of use could be construed as ‘contracting out’ of the proposed exception, therefore rendering it unenforceable. It is also uncertain what effect the proposed exception will have on the deployment of the Robots Exclusion Protocol, under which text files can be placed on a website to communicate with web crawlers and set the parameters of access to that website.

The proposed exception is also likely to be opposed by some right holders. Copyright exceptions and limitations are required to pass the ‘three-step test’ (found in the Berne Convention, the TRIPS Agreement, and the WIPO Copyright Treaty). This means that copyright exceptions/limitations are only permitted in (1) certain special cases (2) which do not conflict with a normal exploitation of the work, and (3) do not unreasonably prejudice the legitimate interests of the right holder. It is open to question whether what appears to be a broad exception is a “special case” and also the extent to which there is conflict and unreasonable prejudice to right holders. 


artificial intelligence, copyright, database rights, technology, value in data, health tech