As artificial intelligence (AI) increasingly contributes to the creation of music, art and literature, copyright laws—originally devised to protect human creativity—must evolve to tackle the novel challenges posed by AI‑generated works. This shift involves two principal domains: recognising a human author for AI‑generated content and managing copyright infringements involving AI.

Recognising a Human Author for AI‑Generated Works

Copyright law traditionally rewards individuals for their original intellectual endeavours, premised on human creativity. However, AI lacks legal personhood and creative intent, rendering it ineligible for authorship under most existing frameworks. To close this gap, legislation must pinpoint a human figure who can be deemed responsible for the AI‑generated output.

The UK model: a statutory solution

The UK Copyright, Designs and Patents Act 1988 (CDPA) offers a statutory route for attributing authorship to computer‑generated works. Specifically:

Section 9(3) provides:

‘In the case of a literary, dramatic, musical or artistic work which is computer‑generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken.’

Section 178 defines ‘computer‑generated’ as:

‘A work generated by computer in circumstances such that there is no human author of the work.’

Together, these provisions allow authorship to vest in the person who makes the AI possible. Potential candidates include:

• The programmer

The individual or team that designs and trains the AI model. For instance, the developers of a music‑generating AI such as OpenAI’s MuseNet could qualify as authors if their technical input crucially shapes the output.

• The user

The person providing inputs, such as selecting parameters or curating datasets. For example, an artist using an AI tool to generate a painting by specifying styles and colours might claim authorship.

• The commissioner

An entity commissioning the AI for creative purposes, such as a film studio employing AI to compose a soundtrack.

Further it should be noted complex AI projects often involve multiple contributors, giving rise to ambiguity. To resolve this:

• Joint authorship

Copyright law could mirror models for collaborative human works (e.g. films or albums), whereby an AI‑generated novel might credit both the developer who trained the model and the writer who shaped the narrative via prompts.

• Significant creative‑contribution test

In disputes, courts could assess who exerted the greatest influence on the final work. For example, if an AI autonomously generates a symphony with minimal human intervention, the programmer’s role might be deemed decisive.

Practical example:

Consider an AI tool such as DALL‑E producing a surrealist painting from the prompt ‘a melting clock in a desert’. Both the programmer who trained the model on art‑history datasets and the artist who supplied the prompt could share authorship, thereby ensuring the work benefits from copyright protection while acknowledging the combined human efforts.

Addressing Copyright Infringements Involving AI

AI can infringe copyright at two principal stages:

• Data processing

Collecting, digitising and extracting information from protected works (e.g. converting books into machine‑readable form) may breach reproduction rights.

• Output similarity

AI‑generated outputs that closely resemble copyrighted material could indicate infringement, particularly if they reveal unauthorised use of protected works.

To tackle these issues while fostering AI innovation, copyright law might adopt a balanced approach that melds the three‑step test from the Berne Convention with fair‑use factors. South Korea’s Copyright Act exemplifies such a hybrid.

South Korea’s hybrid approach

Amended in 2012, the Korean Copyright Act introduced Article 35‑3:

‘A copyrighted work may be used … if such use does not conflict with the normal exploitation of the work and does not unreasonably prejudice the legitimate interests of the copyright holder.’

This provision integrates the three‑step test—ensuring limited use, normal exploitation and minimal prejudice to rights holders—while incorporating four fairness factors akin to those in US law (17 USC §107):

• Purpose and character of the use (e.g. commercial versus non‑profit)
• Nature of the copyrighted work
• Amount and substantiality of the portion used
• Effect on the potential market for the original work

Unlike the United States, South Korea retains specific exceptions (e.g. for quotations or education) alongside this open‑ended fair‑use clause, granting courts the flexibility to apply either framework.

In principle, yes it is possible.

The algorithm is not an author. Period. And it should remain this way. Otherwise, we may end up with a situation in which the owners of artificial intelligence models deprive - through their lobbying power - artists of most of the income generated from AI-supported creativity. One can only imagine some neighbouring rights for human characters reconstructed by AI (apart from public figures, of course).

Legislation in this area should be enacted to preserve the creative nature of human intellectual output, to avoid a reliance on artificial intelligence that undermines the unique qualities of human artistic endeavors. It is important to curb the rampant adoption of artificial intelligence through two methods: 1. Updating legislation to preserve the authenticity of human intellectual output. 2. Disseminating a societal culture that artistic intellectual output generated by artificial intelligence is inauthentic and does not reflect the genius of its creator.

I believe copyright laws need to evolve to address the growing influence of AI in creating music, art, and literature. Currently, these laws are designed to protect works created by human authors, but AI-generated content challenges that framework. I reckon that the law should recognize AI as a tool used by humans, rather than granting copyright to the AI itself. This could mean that the person or entity controlling the AI should hold the copyright, assuming they’ve directed the AI’s creative process in a significant way. However, it’s also important to ensure that AI doesn’t simply exploit pre-existing copyrighted works, so there should be mechanisms to safeguard against this kind of infringement. I believe copyright laws should adapt to both encourage innovation and protect the rights of human creators while balancing the role AI plays in the creative process.