Neural Network - March 2025

In Regulatory and Government updates, the UK Government's consultation on the future legal regime for the use of copyrighted works to train AI models has faced criticism and the European Commission announced that it will be asking European companies to give their views on an EU initiative for a "European Data Union Strategy".

In AI enforcement and litigation news, the ICO became the latest regulator to request information from DeepSeek; Google argued in support of its motion to dismiss an US copyright case over AI training; the European Data Protection Board adopted principles on age assurance and created a task force on AI enforcement; and a major international operation that tackled AI-generated child abuse materials led to 25 arrests.

In Technology developments and market news, we detail the key outcomes of the Paris AI Action Summit.

More details on each of these developments is set out below.

Regulatory and Government updates

• UK Government's AI and copyright proposals meet with backlash
• "European Data Union Strategy" – Companies asked their views

Enforcement and Civil Litigation

• ICO becomes latest regulator to request information from DeepSeek
• EDPB creates a task force on AI enforcement
• Major international operation tackling AI-generated child abuse materials leads to 25 arrests
• Google files reply in support of its motion to dismiss US copyright case over AI training

Technology developments and market news

• Outcomes from the Paris AI Action Summit

Regulatory and Government updates

UK Government's AI and copyright proposals meet with backlash

The UK Government's consultation on the future legal regime for the use of copyrighted works to train AI models has concluded, amid a backlash over its key proposals.

The consultation, which opened in December 2024, proposed to extend the "text and data mining" exemption under the Copyright, Design and Patents Act 1998 to cover commercial use cases, whilst also providing an "opt-out" for copyright holders. We covered the consultation's proposals in more detail in a previous edition of the Neural Network, which you can read here.

These proposals have met with a broadly negative reaction from the UK's creative industries, including a campaign in which practically all the major UK national newspapers prominently displayed the same slogan, "MAKE IT FAIR", on their front pages on 25 February.

Parliamentarians have also signalled their dissatisfaction with the proposals. In the course of the Data (Use and Access) Bill's passage through the House of Lords, Baroness Kidron proposed an amendment that would require greater transparency from providers of web-crawling software (of the sort often used by AI model developers to build datasets on which to train their models), in order to give greater opportunities for copyright holders to enforce their rights. Baroness Kidron's amendment was successfully adopted in the House of Lords – it remains to be seen whether the Government will seek to undo the amendment, now that the Bill has moved over to the House of Commons.

The Government has not yet given a timeframe within which it expects to publish its conclusions from the consultation.

"European Data Union Strategy" – Companies asked their views

On 27 February, the European Commission (the "Commission") announced that it will be asking European companies to give their views on an EU initiative for a "European Data Union Strategy".

The aim of the initiative is to streamline the existing EU data rules, with the public consultation seeking views on how to create "a simplified, clearer and more coherent legal framework for businesses and administrations to share data more seamlessly and at scale, while still upholding high privacy and security standards".

Additionally, the Commission has said that the initiative "may explore options to increase the availability of high-quality data, consider possible investments in data infrastructures, examine ways to use data and reduce the administrative burden and address the external aspects of data flows".

This initiative follows criticism from tech companies that the EU has introduced various new rules recently, which have overburdened them with requirements. Last year, several companies from the EU and US criticised the EU's regulations, stating that the EU risks becoming an "artificial intelligence backwater" as a result of "fragmented and unpredictable" rules.

The public consultation shall begin by the end of this month, with the initiative planned for the end of September of this year.

Enforcement and Civil Litigation

ICO becomes latest regulator to request information from DeepSeek

The Information Commissioner's Office ("ICO") has become the latest in a series of national data protection authorities to request further information from AI model developer DeepSeek. DeepSeek is an AI development startup based in China who's capabilities rival technology developed by OpenAI, the maker of ChatGPT.

Following DeepSeek's explosive arrival onto the market in January 2025, several EU regulators, as well as other data protection authorities worldwide, rapidly initiated some form of regulatory intervention. These ranged from requests for information to full bans on DeepSeek's model, as was implemented by the Italian Data Protection Authority, the Garante, in February.

We covered the range of regulatory action already underway against DeepSeek in a previous edition of Neural Network, which you can read here. The ICO has now also written to DeepSeek to "request information on its approach to data protection for UK residents".

After initially saying, in a response to the Garante, that it did not consider itself to be subject to the EU GDPR, DeepSeek has also now updated its privacy notice in an apparent attempt to comply with the EU GDPR. We are not aware of any indication to date of its position on the UK GDPR.

EDPB creates a task force on AI enforcement

As discussed in last month's Data Protection Update (see here for a link), the European Data Protection Board ("EDPB") has decided to extend the scope of its ChatGPT task force to consider AI enforcement. Speaking on this decision, the chair of the EDPB, Anu Talus said: "The GDPR is a legal framework that promotes responsible innovation. The GDPR has been designed to maintain high data protection standards while fully leveraging on the potential of innovation, such as AI, to benefit our economy. The EDPB's task force on AI enforcement and the future quick response team will play a crucial role in ensuring this balance, coordinating the DPA's actions and supporting them in navigating the complexities of AI, while upholding strong data protection principles". The task force is expected to coordinate the approach the EU authorities take to data protection law enforcement in the context of AI models.

Major international operation tackling AI-generated child abuse materials leads to 25 arrests

Europol, the EU law enforcement agency, has announced that at least 25 arrests have been made worldwide following an international operation targeting individuals who have been distributing AI-generated child sexual abuse material ("CSAM").

The arrests were made on 26 February, in an operation led by Danish law enforcement and involving at least 18 other national authorities, including in the UK. The main suspect, a Danish national arrested in November last year, is alleged to have run an online platform for sharing material of this sort. The operation, named "Operation Cumberland", is ongoing and may lead to further arrests.

AI-generated CSAM has recently become a prominent focus for both law enforcement and legislators. The UK recently announced proposals to legislate to create a new offence of possessing, creating or distributing AI tools designed to generate CSAM, and the EU is also currently discussing proposals for a new regulation to tackle the issue.

Google files reply in support of its motion to dismiss US copyright case over AI training

Last month, Google filed a reply in support of its motion to dismiss the US class action against it, which was also filed last month. The class action accuses Google of using web-scraped data in order to develop its artificial intelligence products. Compensation is sought for the artists and authors whose registered copyrighted works were used, without consent, by Google to develop its generative artificial intelligence models.

Technology developments and market news

Outcomes from the Paris AI Action Summit

The AI Action Summit, held in Paris on 10 and 11 February 2025, has concluded with a number of notable outcomes.

The summit saw the publication of a new International AI Safety Report, with contributions from 96 AI experts, supervised by an expert advisory panel comprised of representatives from 30 national governments, the EU, the Organisation for Economic Co-operation and Development ("OECD"), and the United Nations. The report summarises scientific evidence on three "core" questions:

• What can general-purpose AI do?
• What are risks associated with general-purpose AI?
• What mitigation techniques are available to manage these risks?

Also announced at the summit were the launch of a $400m fund to support the development of AI in the public interest, known as "Current AI" and the formation of the "Coalition for Sustainable Artificial Intelligence", with a mandate to advance initiatives to ensure AI is supportive of sustainability goals and is itself developed in a sustainable, environmentally-compatible way.

At the summit, France also proposed a declaration on the development of AI technologies, entitled the "Statement on Inclusive and Sustainable Artificial Intelligence for People and the Planet". Most countries in attendance at the summit signed the declaration, which sets out priorities including promoting accessible AI with a view to narrowing digital inequality, ensuring AI systems are ethical and trustworthy, ensuring the AI market does not become concentrated in the hands of too few players, and sustainability. However, the US did not sign the statement due to the opportunities that AI presents, and the UK did not sign the statement amid concerns about national security and "global governance".

Our recent AI publications

If you enjoyed these articles, you may be interested in our podcast series covering various aspects of the evolving world of AI and its implications for businesses and broader society. The full podcast series is available for listening here.

You might also be interested in our ongoing article series on the Data (Use and Access) Bill. This month we have published articles on the Bill's proposals for new smart data schemes and on the new statutory register for digital ID verification services that the Bill proposes to create. You can access the full article series here.