UK launches review on AI training using copyrighted content

The UK government has launched a consultation aimed at addressing the growing tension between AI developers and the creative industries over copyright use.

The move seeks to provide clarity on how intellectual property (IP) is acquired and used by AI firms to train their models, ensuring a balance between technological innovation and creator rights.

Amid growing legal battles, including lawsuits involving OpenAI and Stability AI, the UK is positioning itself as a leader in establishing clear copyright standards for artificial intelligence.

UK proposes measures to regulate AI training datasets

The consultation explores multiple proposals to regulate the use of copyrighted material by AI firms.

One key consideration is allowing exceptions to copyright law for AI model training, particularly for commercial purposes.

The proposals ensure that rights holders retain control by reserving their rights.

This provision would enable creators to dictate how their content is used, offering a path for fair licensing and remuneration.

In addition to protecting content creators, the government aims to offer AI developers clarity on permissible materials for model training.

The move addresses mounting concerns from artists, publishers, and content platforms whose work has been scraped to develop generative AI systems such as OpenAI’s ChatGPT and Stability AI’s Stable Diffusion.

Content transparency may become mandatory for AI developers

Another significant proposal focuses on increasing transparency regarding training datasets.

AI firms could be required to disclose the origin and nature of datasets used to train their models, ensuring that creators understand when their work has been utilised.

This measure directly addresses the concerns of intellectual property holders who argue that content is being used without consent or compensation.

This proposal could ignite controversy.

Tech companies remain guarded about revealing their dataset sources, citing commercial sensitivities and the risk of competitors replicating their models.

Given the commercial value of generative AI, such transparency requirements could meet resistance from AI developers keen to protect their trade secrets.

Legal battles highlight UK need for AI copyright clarity

The consultation comes at a critical time when legal disputes over AI copyright usage are intensifying globally.

Last year, The New York Times filed a lawsuit against Microsoft and OpenAI for allegedly infringing copyright by using its content to train large language models.

In the UK, Getty Images sued Stability AI for scraping millions of images without consent, a case that continues to raise questions about IP compliance for AI systems.

The UK’s efforts stand in contrast to the US, where tech lobbying often complicates legislative progress.

Industry analysts believe the UK is better positioned to prioritise personal intellectual property rights.

Under previous leadership, the government attempted a voluntary AI copyright code of practice but struggled to secure broad adoption.

Balancing creator rights and technological advancement

The proposed measures aim to balance the interests of creators and AI developers, fostering innovation while ensuring rights holders are compensated.

To achieve this, the government is encouraging collaboration between the creative and tech industries to establish widely adopted standards for rights reservation and licensing.

The growing shift towards “multimodal” AI, which integrates text, images, and video, adds urgency to the issue.

Recent developments, such as OpenAI’s AI video generation tool Sora, underscore the need for comprehensive regulations to address the evolving capabilities of AI systems.

Sora’s release highlights how AI models are expanding beyond text-based applications, amplifying the need for clarity in IP usage.

The UK’s consultation signals its intent to become a global leader in AI copyright governance.

By prioritising transparency and creator compensation, the government aims to create a framework that supports technological growth while protecting intellectual property.