When Elton John accuses the government of “committing theft”, it’s clear the battle over artificial intelligence and copyright has hit a high note. He’s among a growing number of prominent artists calling for greater transparency over how AI companies are using copyrighted material, and the controversy has peers and ministers locked in legislative “ping-pong” over proposed protections in the Data (Use and Access) Bill.
This debate is more than a celebrity soundbite. The Culture, Media and Sport Committee inquiry into large language models and generative AI, echoing many industry voices, has warned that failing to protect UK creatives risked damaging one of the country’s most valuable sectors, worth £124 billion a year.
For businesses working in publishing, design, marketing, fashion, tech or media, AI is already raising difficult questions around copyright, licensing and originality. And while formal reforms are still being shaped, the direction of travel is clear: companies will need to understand how their creative output might be used to train AI and how they can respond.
While the government has dropped its earlier proposal for a broad copyright exception for AI learning, a move that had alarmed rights holders, it is now leaning towards a voluntary code of practice to govern how AI developers use copyrighted material in training datasets.
The proposed reforms include a new ‘text and data mining’ exception that would allow AI models to use copyrighted materials for training, unless the creator has explicitly opted out. While the government says the aim is to build trust, promote transparency and support innovation, concerns remain that creators could lose control over how their work is used.
This is particularly relevant for businesses creating digital content, images, branding or written materials, where AI tools may learn from publicly available data - including websites, product descriptions or marketing collateral - to generate new content.
For businesses in the creative sector the implications are serious. While the government is suggesting opt-out mechanisms such as metadata instructions or ‘do-not-train’ lists, these could be difficult to manage in practice, particularly for smaller businesses.
Organisations also need to be aware of the risk of inadvertently using AI-generated content that may be based on copyrighted source material. Without clear licensing frameworks or transparency obligations on AI developers, this raises the risk of legal action down the line and potential reputational damage.
Explained commercial law expert Julian Milan of mfg Solicitors: “Now’s the time for businesses to really take stock. Those creating original content should review how they protect and label their work and stay alert to developments around opt-out systems.
“And those using generative AI in their operations should take extra care to document sources, use only approved tools, and have clear policies in place to guard against infringement.”
Julian added: “In this rapidly evolving area, businesses should not assume that silence equals safety. If you’re creating, licensing or using content that might intersect with AI systems, now is the time to get your house in order.”
Full details of the consultation and proposed reforms is available via the government website here
Comments