The UK Government launches a consultation on a text-and-data mining copyright exception
Key Contact: Declan Goodwin
As artificial intelligence continues to revolutionise countless sectors, there are tensions between the rights of content creators and the ambitions of AI developers. At the heart of this conflict lies the use of copyrighted works for AI training and the legal framework governing text-and-data mining (TDM), now to be explored in a government consultation.
The battle between IP owners and AI developers
Content on the internet is often used to train AI systems to create new content or improve machine learning models. However, rightsholders contend that this use often occurs without proper compensation or acknowledgment, potentially undermining their economic and creative interests.
On the other hand, AI developers argue that access to vast datasets, including copyrighted material, is crucial for innovation. Restricting this access, they claim, could stifle technological progress and diminish the global competitiveness of AI research and development.
The tension between these perspectives has led to a critical policy discussion: how can the rights of copyright holders be balanced with the need to foster AI innovation?
UK Government’s proposed solution
The UK Government has launched a consultation to explore new frameworks for text-and-data mining. The proposed approach aims to align the UK more closely with the European Union’s copyright laws while attempting to accommodate the needs of AI developers.
Under the government’s proposed TDM exception, copyrighted works could be used for AI training by default unless the copyright holder explicitly opts out. This approach marks a shift from the current framework and seeks to strike a balance between accessibility and control.
What would this look like?
If implemented, the TDM exception would allow AI developers to mine datasets, including those containing copyrighted material, without seeking prior permission. This would significantly reduce barriers for AI researchers and developers, enabling faster progress and more innovative outcomes.
For copyright holders, the system would establish a mechanism for reserving their rights. Instead of requiring proactive consent for TDM, they would need to take specific steps to opt out and prevent their works from being included in datasets used for AI training.
By adopting a similar approach to the EU, the UK hopes to create a more consistent regulatory landscape while fostering innovation within the AI sector.
The consultation explains how the default permission and opt-out process could work, as well as the differing rules that could apply to commercial and non-commercial use.
Visit here for full details on the consultation and how to respond.
For advice on exploiting and protecting your IP portfolio, contact our Intellectual Property team.
