UK government AI report confirms decision on protection of copyright works

The government's response highlights that the use of copyright works as AI training data will infringe copyright unless permitted under licence or an exemption.

The response also notes that the government has abandoned plans to extend the text and data mining exception to cover the development of AI for commercial use. This exception allows researchers to make copies of any copyright material for the purpose of computational analysis if they already have the right to access the work.

The response also shows the government’s commitment to a code of practice on copyright and AI, stating that those involved in the development of the code are currently “highly engaged” in the process and gives a new date for publication of the code as early 2024.

The planned new code of practice aims to balance the rights of content creators with the desire of AI developers to train their systems using quality data. The code is currently being developed by a working group established by the UK’s Intellectual Property Office (IPO), which consists of representatives from the technology, creative and research sectors. The IPO said the code will aim to “make licences for data mining more available” and help to “overcome barriers that AI firms and users currently face and ensure there are protections for rights holders”.

Previously, government proposals to amend legislation enabling greater use of copyright works to train AI were met with backlash from the creative sector. However, the latest government response “makes it clear that its current approach on this issue will fully respect the rights of content creators, despite the development of AI technologies remaining a priority”, said Gill Dennis, an expert in IP law at Pinsent Masons.

Although informal comments were made in various parliamentary committees last year that the government had abandoned its plans to extend the text and data mining exception, this is the first official confirmation we have had of that policy, with Dennis saying this will “come as a relief to copyright owners”.

The publication of the Code of Practice was originally due in the summer of 2023, before being pushed back to autumn 2023. However, it has now been moved to early 2024, with Dennis adding: “This reflects the difficulty of achieving a consensus on use of copyright works as AI training data between stakeholders with directly opposing interests on this issue”.