Cultural industries must be protected from ‘danger’ of generative AI, says report

New measures are needed to protect the cultural industries from the rise of generative AI and avoid a scenario where UK creators are sacrificed in favour of the large-scale, unlicensed use of creative content by models trained overseas.

That’s the stark warning in a report produced by the House of Lords Communications and Digital Committee. It calls for “a licensing-first regime, underpinned by robust transparency, that safeguards creators’ livelihoods while supporting sustainable AI growth”.

Such a model would ensure AI developers obtain permission and pay fair remuneration to rightsholders. The committee says that the UK is well-placed to benefit from the emerging market for licensing content for AI.

“The government should support this market to grow in a way that works for AI developers and rightsholders of different sizes, rather than relying on a single initiative such as the Creative Content Exchange,” the report says.

Recommendations include:

• Developing sovereign AI models that deliver enhanced transparency and respect for copyright as a default.
• Making transparency about AI training data a statutory obligation, introducing a mandatory transparency framework for UK AI developers and investigating how public procurement and regulatory tools could promote compliance by international developers.
• Introducing protections against unauthorised digital replicas to give creators and performers clear control over commercial exploitation of their identity.

The government has already dropped its support for a new commercial text and data exception with an opt-out mechanism for training commercial AI models, but the committee calls on the UK to follow in the footsteps of Australia and rule this out altogether.

“Our creative industries face a clear and present danger from uncredited and unremunerated use of copyrighted material to train AI models,” said the committee’s chair, Baroness Keeley.

“The government’s task should be to create the conditions that will allow a licensing-first approach to AI training to flourish, backed by effective transparency requirements and technical standards for data provenance and labelling, so that rightsholders and developers can participate confidently in this emerging market.

“The future for AI in the UK should be based on transparent and responsible use of training data. We are calling on the government to embrace the opportunities this presents, and to demonstrate its commitment to the UK’s gold-standard copyright regime and our outstanding creative industries in its forthcoming economic assessment and update on AI and copyright.”

The Creative Content Exchange

Earlier this year, 12 cultural organisations (including the National Portrait Gallery and Historic England) were confirmed as participants in a pilot digital marketplace known as the Creative Content Exchange where digitised cultural and creative assets can be purchased by AI developers.

As part of its work, the committee heard a wide range of views on the platform. The report states that there was strong consensus among creative-industry representatives that the exchange must not “displace or undermine established licensing models”. There are also concerns that it could be used to justify or facilitate a “de facto opt-out regime for AI training”.

“Given the UK’s wealth of high-quality creative and informational data, AI firms will have strong incentives to negotiate for licensed access if government policy creates clear incentives for a licensing-first approach, supported by effective transparency requirements and enforcement, rather than tolerating large-scale unlicensed training as a competitive advantage,” the report says.

“The Creative Content Exchange is a welcome pilot to test how cultural and heritage datasets can be licensed and used for AI and other data driven purposes. However, it cannot by itself resolve underlying questions about compensation for the use of copyrighted works in AI training. Its value will depend on forming part of a wider framework that upholds lawful licensed use, strengthens transparency and enforcement, and ensures fair remuneration for rightsholders.”