YOUR GO-TO INTELLIGENCE RESOURCE FOR THE SCHOLARLY PUBLISHING INDUSTRY

Cambridge University Press & Assessment has pursued licensing arrangements intended to protect creative industries as governments worldwide consult on the use of copyrighted works to train Large Language Models (LLMs). Over the past 12 months, the organization has submitted consultation responses to the Indian government, the European Union (EU), and the UK government as part of ongoing policy discussions.

Cambridge, alongside other organizations including the Publishers Association in the UK, has called for equitable use of copyrighted data by AI models. These positions include transparency regarding copyrighted content used in LLM training, fair remuneration for that use, and clear attribution of source material to ensure copyright owners receive recognition, which is described as supporting trust in the validity of AI-generated content.

In February 2025, Cambridge responded to a UK government proposal that could allow AI companies unrestricted access to copyrighted works unless rights holders proactively opted out. The response highlighted risks associated with this approach, including the potential for copyrighted material such as academic research and learning resources to be scrapped without sufficient protection, and noted that proven licensing solutions are already in place and in use by some AI companies.

More recently, consultation responses were submitted to the European Union (EU) and to the Government of India’s Working Paper on Generative AI and Copyright. The EU consultation focused on the practical implementation of the 2024 EU Artificial Intelligence Act, including existing systems that allow copyright owners to opt out of their works being used to train LLMs. Cambridge’s response questioned the effectiveness of those systems, describing them as insufficiently robust to ensure copyrighted works are excluded from AI training.

In India, the government is consulting on proposals intended to balance AI innovation with copyright protections, including a “Hybrid Model” that would provide blanket access to copyrighted content through a single flat fee, with copyright owners receiving a share of the revenue. While the stated aims include simplifying licensing, reducing transaction costs, and ensuring fair access to training content, Cambridge argued that these aims would not be achieved and that the model would undermine foundational principles of copyright law, potentially harming creative industries in India and elsewhere.

The response instead supported the adoption of existing collective and individual licensing solutions described as already proven effective and aligned with India’s “Orange Economy,” a model adopted in 2025 that focuses on supporting the country’s creative industries. The announcement also emphasized that ethical and responsible AI development depends on creators and rights holders being recognized and compensated when their works are used in AI training, and that copyright frameworks already provide meaningful safeguards in many markets.