UK AI Copyright Code Burdened With Disagreements

Introduction              

Artificial intelligence (AI) models have sent shockwaves through the creative industry, raising important questions on the copyright implications that incur when training these models. While AI companies and creative industries are awaiting guidance, the UK’s highly anticipated code setting out the rules on training AI models has been shelved. How did we get to this point and what is coming next?

The Need for a Code

The way AI models are trained and the content they put out has raised a number of legal issues, namely in the realm of copyright law. For example, this article explores how record labels, including Universal Music, are suing AI company Anthropic for breach of their copyrights.

There are concerns that AI models are reproducing authors’ and musicians’ copyrighted work without payment. Indeed, as these models scan the internet to process large amounts of data, it is difficult to understand quite how much content is being used for training, and whether this content may be protected.

Therefore, the Intellectual Property Office (IPO), the UK government’s agency for overseeing copyright law, has been in talks with the creative industries and AI companies in order to produce guidance. However, while the former believes AI companies need permission to train their models on copyrighted content, the latter argue this is covered by copyright exceptions related to data mining.

Why was the Code shelved?

These two opposing sides are keen to defend their interests. However, this has led to a breakdown in discussions. The group of industry executives convened by the IPO, from both sides of the argument, have been unable to agree on a voluntary code of practice.

Therefore, the responsibility has returned to officials at the Department for Science Innovation and Technology.

What Happens Next?

In the wake of the code of practice being shelved, alternative arrangements will need to be made in order to provide clarity in this area.

It is expected that the government will publish a white paper in the coming days that will set out a number of suggestions for further regulation of AI in the UK. Amongst the proposals, this paper is likely to discuss arrangements for AI and copyright.

However, there is concern that, while these proposals may provide clarity, they will fall short of any definitive policy suggestions. As a fast-moving area, which contributes to growing conflict between AI companies and the creative industries, clear policy or even legislation may be needed to alleviate tensions.

Conclusion

The intersection of artificial intelligence and copyright law has prompted significant debate and uncertainty within the creative industry. Despite efforts by the UK government to establish guidance through a code of practice, disagreements between AI companies and creative stakeholders have led to its shelving. As the discussion moves forward, forthcoming proposals for AI regulation may offer some clarity, but concerns remain regarding the need for concrete policy or legislation to address the complex issues surrounding AI and copyright.

By Scott Hickman