Copyright and Artificial Intelligence: Navigating Human Authorship in the MENA Region and Beyond

6 min 14 sec December 3, 2025 (Edited)

AI and the Changing Copyright Landscape

Artificial intelligence (AI) is transforming the creative landscape at an unprecedented pace. From music and literature to visual arts and software development, AI-powered tools can now generate content that rivals human creativity. While this technological evolution presents exciting opportunities, it also raises significant legal and regulatory challenges, particularly in the realm of copyright law.

The United States Copyright Office has been at the forefront of addressing these challenges. On May 9 2025, it released the third instalment of its AI initiative — a  report examining the legal implications of training generative AI models using copyrighted materials. This report forms part of the Office’s broader effort to explore how copyright principles apply in the age of AI, including issues around data usage, authorship, and registration of AI-assisted works.

Under current US practice, any work containing AI-generated material must clearly disclose human involvement and satisfy the requirement of human authorship, which remains the cornerstone of US copyright law. This requirement raises complex questions when AI acts as the primary creator of a work.

In the Middle East and North Africa (MENA) region, the approach is less developed, with legislation often pre-dating the AI revolution. The UAE Copyright Law, for instance, recognises only natural or legal persons as authors. Consequently, works created solely by AI may not qualify for copyright protection under existing frameworks.

This article examines the current legal landscape in the UAE and the broader MENA region, highlights challenges presented by AI, and explores practical approaches for creators and businesses seeking to navigate emerging copyright issues.

Understanding AI as a Creative Tool

AI is increasingly being used as a creative tool rather than a mere utility. Modern AI systems can compose music, write poetry, design digital art, and even generate software code. These systems operate by learning patterns from massive datasets and producing outputs based on algorithmic inference. A key distinction arises between AI-assisted creation — where a human guides or supervises the output — and fully AI-generated works, where the machine independently produces the content.

The implications of AI-generated content are far reaching. In the music industry, for example, an AI-generated track mimicking the vocal style of artists such as Drake or The Weeknd went viral, sparking intense public discussion. Listeners were shocked to learn that the music they were hearing was not performed by the original artists but by an AI system trained to emulate their voices. This incident illustrates the core tension in copyright law: traditional frameworks rely on the notion of human authorship, yet AI is increasingly capable of producing works that are indistinguishable from human-generated creations.

The situation raises real questions for creators, companies, and regulators. Should content made by AI be protected by copyright at all? And, if so, who would own those rights — the person who built the AI, the person who used it, or maybe no one at all? These are important questions that show why clearer rules are needed, especially in places like the UAE, where copyright laws still assume that only humans can be authors.

Human Authorship Requirement in Copyright Law

Human authorship is a foundational principle in copyright law. Both in the US and UAE, copyright protection is granted to works created by human beings or, in some cases, recognised legal entities. In the US, the Copyright Office explicitly states that works produced entirely by a machine, without human intervention, cannot be registered for copyright. Similarly, the UAE Copyright Law recognises only natural or legal persons as authors, which inherently excludes AI as an author.

This principle ensures that copyright serves its dual purpose: incentivising human creativity and granting enforceable rights to authors. However, AI challenges this traditional framework by blurring the lines of creative responsibility. When an AI system generates a work with minimal human input, it becomes difficult to determine whether copyright protection should attach, and if so, to whom.

The human authorship requirement also has implications for commercial and artistic industries. For example, if a company releases an AI-generated track or artwork without clear human authorship, it may be ineligible for copyright protection in the UAE or US. This can affect licensing, distribution, and enforcement rights, leaving creators and businesses exposed to legal and financial risks.

Training Data and Emerging Copyright Obligations

Beyond authorship, another critical copyright challenge lies in the use of protected materials for AI training. Generative AI systems rely on vast datasets that may include copyrighted works, raising questions about whether copying and analysing such material without permission constitutes infringement.

Recent legal and policy developments in the EU, UK, and US have begun to define the boundaries of lawful training. The EU’s Copyright in the Digital Single Market (CDSM) Directive introduces two “text-and-data-mining” (TDM) exceptions: a narrow one for scientific research and a broader one allowing commercial use, provided rights holders can opt out in a machine-readable manner. Under the EU Artificial Intelligence Act, which entered into force in August 2025, providers of general-purpose AI models must also maintain a policy to comply with copyright law and publish a summary of their training data sources. These measures signal a clear trend toward transparency, opt-out compliance, and dataset provenance governance.

Although MENA jurisdictions have not yet introduced specific TDM exceptions or transparency obligations, the evolving EU and UK frameworks carry important implications for regional businesses developing or deploying AI systems that interact with these markets. Even if training occurs outside the EU, companies placing AI systems in the EU market, or collaborating with global AI providers, may need to respect opt-outs, maintain data-use records, and disclose dataset summaries.

Establishing internal governance measures, including dataset mapping, opt-out monitoring, and crawler oversight, will help MENA creators and developers align with emerging international standards, reduce compliance risks, and strengthen the credibility of their AI initiatives in a rapidly evolving global legal environment.

Copyright Registration of AI Works in the US

The US Copyright Office has been actively addressing the implications of AI on copyright through its ongoing AI initiative. Part 3 of this initiative, released in May 2025, focuses on the intersection of AI-generated content and data usage. The initiative clarifies that while AI can be a tool in the creative process, the law requires clear disclosure of AI involvement and boundaries of human authorship for a work to qualify for copyright registration.

Under current US rules, any work that contains AI-generated material must explicitly identify the human author(s) responsible for the work. The Copyright Office will not recognise AI as an author, reflecting a fundamental principle: copyright protection attaches to human creativity, not machine output. This policy applies even if AI is the primary source of content generation; without meaningful human input, the work cannot be copyrighted.

A notable example illustrating this challenge is the viral AI-generated music track mimicking the voices of the famous signer, Drake, and the known band, The Weeknd. When released online, listeners were initially unaware that AI had been used to replicate the artists’ vocal styles. The incident sparked widespread debate about whether the resulting work could be considered “authored” under US copyright law. It highlighted the limitations of the current legal framework, which cannot recognise a machine as an author, and emphasised the necessity of human involvement in AI-assisted works to secure copyright protection.

The implications for creators and businesses are significant. Music producers, visual artists, and writers must document their human contributions when using AI tools to generate content. Failure to do so may result in unprotected works, exposing creators to potential misuse, unauthorised copying, and commercial loss.