The Battle for Creative Rights: How UK's AI Copyright Proposal is Reshaping Global Law

The UK government has proposed significant changes to copyright law that could fundamentally alter how AI companies use creative content. The proposals, outlined in the December 2024 "Copyright and Artificial Intelligence" consultation paper, have sparked intense debate among creators, tech companies, and legal experts worldwide. This article examines the UK's approach, compares it with other jurisdictions, and considers the broader implications for creative industries.

The UK Proposal: A Controversial Exception

The UK government's preferred approach (Option 3 in their consultation) would create an exception for "text and data mining" that would allow AI developers to use copyrighted works without permission, provided:

• The AI developer has lawful access to the works

• The rights holder has not explicitly "reserved their rights" (opted out)

• There is increased transparency about the sources used in AI training

This approach effectively inverts traditional copyright principles by shifting from an "opt-in" to an "opt-out" model. Rather than requiring permission before using creative works, AI companies would be able to use any content unless explicitly forbidden.

The Creative Community's Response

The proposal has triggered widespread opposition from creative communities:

• Over 48,000 creators, including Kate Bush, Abba's Björn Ulvaeus, and Paul McCartney, signed a statement warning of a "major, unjust threat" to artists' livelihoods;

  
  

• More than 1,000 musicians released a silent album titled "Is This What We Want?" to protest what they call the "legalization of music theft"; and

  
  

• The Make It Fair campaign launched nationwide advertising, urging the public to contact MPs

Critics argue that the opt-out model is fundamentally flawed and unworkable because:

• It places an impossible administrative burden on creators to monitor and protect their work across the internet

  
  

• Many creators, especially emerging artists, may be unaware of the need to opt out

  
  

• Technical limitations mean no opt-out system can reliably catch every use of copyrighted work

As Ed Newton-Rex, founder of Fairly Trained, bluntly stated: "Opt-out is an illusion."

Global Approaches: A Comparative View

European Union

The EU's AI Act contains provisions similar to the UK proposal, including an opt-out mechanism for text and data mining. However, recent criticisms from Axel Voss (an architect of the EU's 2019 copyright directive) suggest the approach has created a "devastating" legal gap. Voss argues that the text and data mining exception was intended for limited private use, not to allow large companies to harvest vast amounts of intellectual property.

United States

The US approach relies heavily on the "fair use" doctrine rather than specific AI exceptions. This is currently being tested in numerous lawsuits, most notably The New York Times v. OpenAI and Microsoft. Legal experts note that fair use is determined case-by-case, considering factors like purpose, nature, amount used, and market impact. This creates uncertainty, as no single ruling can be universally applied.

The US Copyright Office has determined that purely AI-generated works are not copyrightable, though human-authored works incorporating AI-generated material may qualify for protection.

Australia

Australia has taken a more measured approach by establishing the Copyright and Artificial Intelligence Reference Group (CAIRG) in December 2023. This standing group engages stakeholders across sectors to address AI and copyright intersections, with initial focus on copyright material as AI inputs and transparency issues. The Australian government is still in consultation phase rather than proposing specific legislative changes.

Historical Context: AI vs. Previous Disruptions

AI's impact on copyright resembles previous technological disruptions like peer-to-peer file sharing (Napster), digital sampling in music, and social media content sharing. However, AI's scale, speed, and capacity to generate derivative works present unique challenges.

While previous disruptions primarily affected distribution of creative works, AI directly affects creation itself by potentially displacing human creators. And unlike previous technologies where infringement was clearly defined, AI training occupies a legal gray area regarding whether it constitutes copying.

Practical Enforcement Challenges

Enforcing copyright in the AI era faces several practical challenges:

• Training datasets are typically not publicly accessible, making it difficult for creators to know if their work has been used

• Once data has been used to train an AI model, it cannot be removed retroactively

• AI models can be trained in jurisdictions with more permissive laws but deployed globally

• The cross-border nature of AI development creates jurisdictional complexities for enforcement

Industry Impact

The consequences of these legal developments vary across creative sectors:

Books and Publishing:

• Authors express concern about AI training on their works without compensation

• The risk of AI-generated books flooding markets could devalue human creativity

• Publishers are divided between pursuing litigation and striking licensing deals

News Media:

• Major publishers like The New York Times are pursuing litigation

• Others have negotiated licensing agreements with AI companies

• Smaller independent news outlets have less leverage in securing fair compensation

Music:

• Musicians fear AI-generated music could threaten livelihoods, particularly for emerging artists

• The UK music industry contributed £7.6 billion to the economy in 2023, highlighting what's at stake

• Current focus is on protecting both compositions and performances from unauthorized AI use

Key Dates and Milestones

• February 25, 2025: UK consultation on copyright and AI closed

• August 2, 2025: EU AI Act provision requiring AI companies to provide summaries of training data comes into effect

• Ongoing 2025: Multiple US copyright lawsuits are expected to reach appellate courts

• 2025-2026: Potential UK legislation following consultation feedback

Conclusion

The UK's proposed copyright exception represents a pivotal moment in the global response to AI. By inverting traditional copyright principles to favor AI development, it risks undermining the economic foundation of creative industries while potentially accelerating AI adoption.

The battle lines are clearly drawn: AI companies argue their systems cannot function without access to vast datasets that inevitably include copyrighted works, while creators maintain that their legal right to control and be compensated for their work must be preserved.

As legal frameworks continue to evolve across jurisdictions, finding the right balance between technological innovation and protecting human creativity remains the central challenge. The outcome will shape not just copyright law, but the future of creative expression itself.