UK Government responds to the AI Action Plan

In January, the UK government set out its response to the AI Opportunities Action Plan that was developed by tech entrepreneur and Chair of Advanced Research and Invention Agency (ARIA) Matt Clifford CBE. The plan set out recommendations for the government, which largely fell under three key goals:

• Invest in the foundations of AI
• Push hard on cross-economy AI adoption
• Position the UK to be an AI maker, not an AI taker.

The government agreed to take forward all 50 recommendations set out in the plan. In its response, the Prime Minister referred to the government’s desire for Britain to be a country that is “going to make AI breakthroughs and export them to the world” rather than be a country that is “left to buy those breakthroughs by importing them.”  The Prime Minister described the plan as “a plan to make our country an AI superpower.”

The plan is nothing if not ambitious. So, what are the notable recommendations and their implications?

Unlocking data assets in the public and private sector

National Data Library (NDL)

The plan recommends the creation of an NDL – noting the need to “responsibly unlock both public and private data sets to enable innovation by UK startups and researchers and to attract international talent and capital.”  

The UK has several rich publicly owned data sources that would be of great interest to private and public market participants. The NDL would aim to make data that already exists available and strategically shape what data is collected.

The establishment of the NDL, not to mention the corresponding guidelines and best practices for releasing open government data sets that can be used for AI, including the development of effective data structures and data dissemination methods, will be significant in determining perceived best practices in the UK for licensing data sets of diverse types.

The government agreed to this recommendation and is expected to set out further details on the NDL and the government’s wider data access policy by Summer 2025.

Private data sets

Other noteworthy recommendations include actively incentivizing and rewarding researchers and industry to curate and unlock private data sets (Recommendation 12) and establishing a copyright-cleared British media asset training data set, which can be licensed internationally at scale (Recommendation 13). 

Further details about Recommendation 12 are expected when the government publishes greater context on the NDL in a few months. As for Recommendation 13, which was only partially agreed upon, the government’s response stated that the Department for Culture, Media, and Sport and the Department for Science, Innovation, and Technology would engage with partner organizations and industry to consider its potential role. The government has indicated a delivery timeline of Spring 2025 for this engagement.

While there would be benefits to having copyright-cleared British media asset training data, persuading private IP rights owners to license their works as proposed will be no mean feat. It, therefore, remains to be seen whether this is even a realistic prospect.

The text and data mining exception (TDM)

One key element that data privacy and AI professionals will be watching is Recommendation 24. The plan recommends reforming the UK TDM regime so that it is at least as competitive as the EU by introducing a TDM exception to copyright infringement. 

The plan explains:

“The current uncertainty around intellectual property (IP) is hindering innovation and undermining our broader ambitions for AI, as well as the growth of our creative industries. This has gone on too long and needs to be urgently resolved. The EU has moved forward with an approach that is designed to support AI innovation while also enabling rights holders to have control over the use of content they produce. The UK is falling behind.”

The government launched a consultation on a TDM regime for the UK in December 2024, which closed in February 2025. The output of the consultation will be instrumental as a first step in outing the practicalities of the UK’s approach.

There is an opportunity for the UK to distinguish itself from the EU by establishing a more flexible TDM exception than under the current EU regime and to move closer to countries like the USA, Canada, Singapore, or Japan, which have taken more liberal approaches. However, it seems unlikely that any approach will be fully aligned with the US notion of “fair use,” which would represent the most radical change to UK law. The published mandate is to make the UK TDM regime “at least as competitive as the EU.” Will it go further than that?

The government’s endorsement of the recommendations is a positive signal. It shows that the government recognizes that, in addition to the necessary infrastructure and incentives, high-quality data will be needed to fuel both frontier AI progress and high-quality AI applications.

Implementing the plan’s (ambitious) recommendations will not be easy, and the key to achieving the UK government’s goal to be an AI maker, not an AI taker, and not falling behind our global competitors, will be aligning the needs of the tech/AI developers sector and the creative industries/other rights holders.

← Back to News