UK plans to let AI companies use copyrighted work freely to train AI models under new copyright exemption

The UK government has floated a controversial proposal to allow tech firms to use copyrighted works for training artificial intelligence (AI) models under a new copyright exemption. This move, unveiled during a consultation on Tuesday, aims to bridge the growing divide between AI developers and creative industries but has drawn sharp criticism from artists, publishers, and rights campaigners.

The proposed changes would permit tech companies would be permitted to use copyrighted material freely, and without seeking the IP holder’s permission to train their AI models, unless the creators of the IP have explicitly opted out of such arrangements. Advocates for creatives argue this undermines their rights and livelihoods, especially in a sector valued at 126 billion British pounds annually. Critics, including book publishers, have slammed the proposal as “untested and unevidenced,” warning it could disproportionately harm smaller creators who may lack the resources to reserve their rights.

Balancing AI innovation with creatives’ IPs

The government argues the proposal could resolve the ongoing standoff between AI developers and creative professionals. It emphasised the need for transparency from AI firms regarding how training data is sourced, used, and the content generated. The Data Protection Minister described the plan as a “win-win,” suggesting it could lead to more licensing opportunities, providing creators with potential new revenue streams.

However, campaigners remain sceptical. They argue that the “rights reservation” mechanism – where creators must actively opt out – is unfair and may only benefit major rights holders while leaving smaller artists exposed. Leading voices in the publishing and news industries also raised concerns, urging the government to prioritise transparency and enforce existing copyright frameworks rather than introduce what they view as “unworkable” opt-out systems.

More from Tech

How ChatGPT is becoming everyone’s BFF and why that’s dangerous America ready for self-driving cars, but it has a legal problem

Creators fight back against AI exploitation

The backlash from the creative community has been robust. Over 37,000 artists, writers, and performers, including high-profile names like Radiohead’s Thom Yorke and actor Julianne Moore, have signed a statement decrying the unlicensed use of creative works for AI training as a major threat to their livelihoods. Campaigners warn the new exemption could pave the way for widespread exploitation, allowing AI firms to sidestep fair compensation.

Some also question whether the proposal would apply to existing AI models like ChatGPT and Google’s Gemini, which have already been trained on massive datasets. This ambiguity, critics argue, further complicates an already fraught situation.

Exploring protections for public figures

The proposal is also exploring broader protections, including a potential US-style “right of personality.” This would safeguard public figures from having their voice or likeness replicated by AI without permission. The debate has gained urgency following high-profile cases like Scarlett Johansson’s dispute with OpenAI, where her voice was allegedly imitated by a voice assistant.

As the government gathers feedback, the proposal remains a lightning rod for debate, with its impact on both AI innovation and the UK’s creative economy hanging in the balance.