More than 400 prominent British artists have raised their voices to demand greater safeguards in the protection of their copyright against the use of protected works in the training of artificial intelligence models.
In a letter to Prime Minister Keir Starmer, figures including Sir Elton John, Dua Lipa, Sir Paul McCartney, Coldplay and Ian McKellen back an amendment to the Data (Use and Access) Bill that would force AI companies to reveal what content they use in their data sets.
The key vote in the House of Lords is scheduled for May 12, 2025, and marks a decisive moment in the balance between technological innovation and respect for creative heritage.
The artists’ letter and the Data (Use and Access) Bill
The letter brings together the signatures of more than 400 creators (musicians, actors and writers) who underline their moral authority over their works.
Artists such as Dua Lipa, Sir Elton John, Sir Paul McCartney, Kate Bush, Robbie Williams and Sir Ian McKellen raise awareness of the need to preserve the cultural and economic value of artistic creation.
The purpose is not to stop the advancement of AI, but to ensure that its development respects the copyright framework and guarantees fair compensation.
The initial text of Data (Use and Access) Bill contemplated an “opt-out system: AI companies could train their models with protected works, unless the authors expressed their opposition beforehand.
The amendment promoted by Baroness Beeban Kidron proposes, instead, a passive “opt-in” mechanism, where developers must register and detail the protected content they use.
This change seeks to facilitate license negotiation and reinforce accountability, creating a dynamic and transparent licensing market.
Parliamentary debate and government position
The Executive of Keir Starmer has shown reluctance to incorporate substantial modifications, arguing the need to balance the protection of creators with the drive for innovation and competitiveness of British AI companies.
The vote in the House of Lords on May 12 will define whether the transparency requirement is incorporated or if the opt-out model prevails, considered insufficient by numerous creators and associations.
Parallel movements in the United States
In May 2025, the US Copyright Office published a report questioning the automatic application of the fair use doctrine to AI training with protected works.
The document warns that the commercial use of said materials could exceed the limits of legitimate use, generating legal uncertainty for the creators.
Legislative initiatives
In the North American Congress, the “Generative AI Copyright Disclosure Act” proposes that AI companies notify one month in advance the Copyright registration of the works used in their training processes.
For its part, the “TRAIN Act” seeks to grant creators access to training records, allowing them to verify the inclusion of their works and negotiate licenses or compensation.
These initiatives reflect a growing interest in establishing clear standards of transparency and accountability.
Transparency and regulation in the European Union
The European Artificial Intelligence Regulation, in force from the beginning of 2025, requires providers of generative models to publish detailed summaries of training data, including copyrighted material.
Although the AI Act does not expressly classify generative systems as “high risk,” it does establish transparency obligations to clarify the origin and nature of the data used.
Criticisms and gaps
Despite its progress, the AI Act has received criticism for the ambiguity in the definition of the text and data mining (TDM) exception, which could allow the massive exploitation of content without adequate compensation to its creators.
Furthermore, the associated Code of Practice has been criticized as weak for failing to ensure effectively binding licensing mechanisms.
Artistic manifestations and litigation
Silent Protest Album
In February 2025, a collective of a thousand artists released a 47-minute silence album whose track titles form the message: “The British Government must not legalize music theft to benefit AI companies”, which translated would be: “The British Government should not legalize music theft to benefit AI companies.”
Sales went to the Help Musicians organization, as a form of symbolic and practical pressure on the British Parliament.
Lawsuits against generative music platforms
In the US, major labels (Sony, Warner and Universal) filed lawsuits against generative music startups that they allege would have used protected recordings without authorization to train their models.
These litigations highlight the high economic and reputational risk for AI platforms that do not have clear licenses.
Economic and cultural implications
The UK creative sector contributes more than £120 billion a year to the economy. A fair licensing mechanism would not only protect this revenue, but would foster innovative collaborations between AI creators and developers, reducing legal uncertainty and litigation.
In turn, transparency frameworks in the US and EU could incentivize business models based on licensing, strengthening a responsible innovation ecosystem that recognizes the value of human creativity.
The mobilization of more than 400 British creators, together with legislative initiatives in the United States and the European Union, marks a turning point in the regulation of artificial intelligence.
The vote in the House of Lords on May 12, 2025 will define whether the United Kingdom opts for a model of transparency that respects the economic and moral rights of authors or if it continues a system that favors indiscriminate access to protected works.
Likewise, the result will serve as a global reference to balance the protection of creativity with the technological drive.
This post is also available in: