Copyright and Generative AI: An Overview
The recent passage of the Data (Use and Access) Act (the “DUA Act”) in the UK on 19 June 2025 has significant implications for copyright issues related to generative AI. This legislation, which received Royal Assent after being passed by the UK Parliament, aims to reform the General Data Protection Regulation in the UK (the “UK GDPR”) and the Privacy and Electronic Communications Regulations 2003 (“PECR”).
The Legislative Journey
Before its passage, the DUA Act underwent extensive debate in Parliament, particularly concerning the training of AI tools on copyright-protected works. The House of Lords proposed amendments to enhance the rights of copyright holders who fear that their works may be utilized for training AI models without their consent.
This amendment arose in response to concerns expressed by content creators during a contentious government consultation on AI and copyright at the end of 2024. The government had suggested expanding the existing text and data mining exception in copyright law, which would place the onus on copyright owners to opt out if they wished to prevent their works from being used for AI training. Notably, many prominent content creators opposed these proposals, leading to significant public discourse.
Key Provisions and Amendments
The House of Lords’ amendment called for the UK Government to implement regulations that would require organizations using web crawlers and general-purpose AI models to publish transparency statements. These statements would detail the tools employed in web crawling and provide information on how intellectual property owners could reach out regarding the use of their works. Such measures could have empowered content creators to identify which of their works were being utilized for AI training, potentially enabling them to take protective action.
However, the House of Commons rejected this amendment, arguing that legislation should await the completion of the government’s consultation responses. The government expressed concerns that incorporating transparency requirements into the DUA Act would complicate the legislation, necessitating separate legal action. This position was informed by ongoing litigation in the UK High Court, which is expected to yield rulings with profound implications for copyright material usage in AI model training.
Government Commitments
Ultimately, the UK government agreed within nine months of the DUA Act’s Royal Assent to undertake two crucial tasks:
- Publish a report regarding the use of copyright works in the development of AI systems.
- Conduct an economic assessment of the impact of various options on copyright owners and AI system developers or users, as outlined in the Copyright and AI Consultation Paper.
The government is required to provide a progress statement on both initiatives within six months.
Looking Ahead: The Future of AI and Copyright
The training of generative AI models using copyright-protected works has emerged as a crucial topic globally. Ongoing litigation and discussions surrounding regulatory reform are poised to shape the future landscape of copyright law as it pertains to AI in both the UK and internationally.
The latest consultation by the UK government on copyright and AI seeks to safeguard and bolster both the creative industries and the AI sector. It is anticipated that the AI copyright report, along with the economic assessment and government proposals, will be released in Spring 2026. Any new draft laws concerning AI copyright could potentially be approved by Parliament by late 2026.
Conclusion
The evolving relationship between copyright law and generative AI necessitates close attention from stakeholders across various sectors. As legislative developments unfold, the implications for content creators and AI developers will be significant, requiring adaptation and proactive engagement with regulatory frameworks.
