Authored By: NDIOHO ABASI-ANYANGA OFONIME
University of Port Harcourt
Abstract
This article examines the burgeoning conflict between Generative Artificial Intelligence (AI) and the traditional frameworks of Intellectual Property (IP) in Nigeria. As AI systems increasingly produce high-value creative content by “harvesting” vast web-based databases, the legal definition of authorship is being pushed to its breaking point. This paper critiques the Nigerian Copyright Act 2022 and the National Digital Economy and E-Governance Bill 2025/2026, arguing that current legislative efforts focus heavily on digital administration while leaving a “vacuum of ownership” for AI-generated works. Through a doctrinal analysis of the “human effort” requirement, the article proposes a new regulatory bridge to protect Nigerian creators in an era of algorithmic creativity. While the Copyright Act 2022 remains the primary law for intellectual property (IP), the upcoming National Digital Economy and E-Governance Bill 2025/2026 is designed to modernise how digital assets and AI are regulated.1
Introduction
The rapid ascent of Artificial Intelligence (AI) has transitioned from the realm of science fiction to the centre of Nigerian legal discourse. In the creative hubs and the tech corridors, a new “author” has emerged: the algorithm. These systems, powered by Large Language Models (LLMs), do not create in a vacuum; they function by scraping, indexing, and “learning” from massive databases of human labour posted across the digital commons. This creates a profound legal paradox: while the input is often protected intellectual property, the output frequently exists in a state of legal limbo.
The current Nigerian legal landscape is at a crossroads. While the Copyright Act 2022 represents a modernised approach to digital rights, it remains tethered to a “human-centric” philosophy that requires a natural person to expend “some effort” for protection to vest.2 Simultaneously, the National Digital Economy and E-Governance Bill — currently the most defining piece of legislation in the 10th National Assembly — seeks to institutionalise Nigeria’s digital transformation.3 However, a critical gap remains. The Bill focuses on the mechanics of the digital economy — e-governance, digital signatures, and infrastructure — but remains largely silent on the substance of ownership for works generated by autonomous agents.
This article provides a critical analysis of this “Authorship Vacuum.” It explores the intersection where the algorithmic author meets the Nigerian creator, arguing that the failure to bridge the gap between AI regulation and IP law risks turning the Nigerian digital space into a “scrapyard of creativity” where innovation is high, but protection is non-existent.
2. The Genesis of the Limbo: Web-Scraping and the “Scrapyard” Argument
The primary fuel for modern Generative AI is the “scraping” of vast quantities of data from the public web. This process involves automated tools that crawl websites, copying and indexing content — from legal journals to artistic portfolios — into massive training databases. While AI developers often frame this as a benign act of “learning,” from a legal perspective, it is a massive, unauthorised reproduction of protected works.
2.1 The Mechanics of AI Training: Harvesting the Digital Commons
AI models do not simply “view” the internet; they ingest it. To train a Large Language Model (LLM), billions of data points are copied into a server to create a mathematical map of human expression. In Nigeria, where digital content creation is a cornerstone of the burgeoning “Orange Economy,” this harvesting often occurs without the knowledge or consent of the original creator. This creates the “Scrapyard of Creativity”: a situation where the raw materials of Nigerian intellect are used to build commercial tools that may eventually compete with the very creators they were built upon.
2.2 Section 108 and the Unauthorised Digital Copy
The Nigerian Copyright Act 2022 introduced a critical update to the definition of a “copy.” Under Section 108, a copy is now defined as “a reproduction in any form including a digital copy.”4 This expansion is significant. Unlike the repealed 2004 Act, the 2022 Act explicitly recognises that a digital reproduction carries the same legal weight as a physical one.
When an AI scraper makes a digital copy of a Nigerian author’s work to store in a training database, it arguably commits an act of infringement under Section 9 of the Act, which grants authors the exclusive right to “reproduce the work in any material form.”5 Unless the AI company can prove that their scraping falls under the “Fair Dealing” exceptions — such as for non-commercial research or private study6 — the very foundation of their model is built on a series of statutory breaches.
3. The Authorship Vacuum: Deconstructing the “Human Effort” Test
If the input (scraping) is a question of infringement, the output (the generated work) is a question of ownership. This is the heart of the “Legal Limbo.”
3.1 The Statutory Definition of a “Person” under Nigerian Law
Ownership of copyright in Nigeria is fundamentally tied to the status of a “qualified person.” The Act provides that copyright is conferred on works where the author is a citizen of, or is domiciled in, Nigeria. However, the Act fails to provide a definition of “person” that includes non-biological entities. In legal interpretation, where a statute does not define a term, it is given its natural and ordinary meaning.7 In the context of creative rights, this has historically meant a human being or a corporate entity — a juristic person.
3.2 Prompting vs. Execution: Does Prompt Engineering Satisfy Section 2(2)(a)?
For a work to be eligible for copyright in Nigeria, it must satisfy the “Effort Test.” Section 2(2)(a) stipulates that a literary, musical, or artistic work shall not be eligible for copyright unless “some effort has been expended on making the work, to give it an original character.”8
This creates a high hurdle for AI users. Does typing a three-sentence prompt into a generator constitute “some effort”? While “prompt engineering” requires skill, the law suggests that the execution of the work — the actual arrangement of pixels or notes — must be directed by the human. If the AI makes the “creative choices,” the human user is merely a patron, not an author. Consequently, without a clear legislative bridge, these works fall into a vacuum where they are owned by no one and can be exploited by everyone.
4. The 2026 Legislative Landscape: A Review of the National Digital Economy Bill
As Nigeria’s 10th National Assembly moves toward the final passage of the National Digital Economy and E-Governance Bill 2026, the focus has primarily been on the “pipes” of the digital economy — infrastructure, electronic signatures, and the digitalisation of public services.9 While these are commendable steps for ease of doing business, the Bill reveals a significant blind spot regarding the “content” that flows through these pipes, specifically Intellectual Property (IP) created by or used for Artificial Intelligence.
4.1 Objectives of the Bill: E-Governance vs. Intellectual Assets
The Bill’s primary objective is to create a unified legal framework for digital transactions and to institutionalise the National Information Technology Development Agency (NITDA) as the central regulatory authority.10 Under Part IX of the Bill, NITDA is empowered to set standards for “emerging technologies,” including AI and machine learning.11 However, the language remains administrative rather than substantive. It regulates the safety and reliability of the algorithms but fails to define the ownership of the data they consume or the works they produce.
4.2 The “Missing Section”: Why the Bill Ignores Algorithmic Ownership
The Bill operates on a “Supremacy Clause” which states that its provisions prevail over any other law in matters relating to the digital economy.12 This creates a potential conflict with the Copyright Act 2022. If the Digital Economy Bill regulates the “use and management” of digital records but remains silent on IP, it leaves Nigerian creators in a weaker position. There is a pressing need for the Bill to be amended to include “IP Disclosure Obligations,” requiring AI developers to be transparent about the databases used to train their models — a move that would bridge the gap between innovation and the protection of Nigerian creative assets.
5. Comparative Perspectives: The EU AI Act and Global Precedents
To understand the “Legal Limbo” in Nigeria, one must look at how other jurisdictions are currently solving the same crisis. The EU AI Act (Regulation (EU) 2024/1689), which entered into full applicability in phased stages from 2025 to 2026, provides a blueprint for “Data Sovereignty.”13
5.1 Transparency Obligations and Database Sovereignty
The European approach introduces a “Transparency Tier” for General-Purpose AI models. Developers are now legally required to publish a summary of the copyrighted material used for training.14 This is the exact “bridge” Nigeria is missing. By forcing transparency, the law allows creators to see if their web-posted works were “harvested” without permission.
In contrast, the United States remains locked in litigation. Cases like Getty Images (US), Inc. v. Stability AI, Inc.15 highlight the “Scrapyard” argument: that using copyrighted databases to build a commercial AI is not “Fair Use” but a direct market replacement. For Nigeria, following the EU’s transparency model within the Digital Economy Bill would offer a more proactive shield for local creators than waiting for years of uncertain court rulings.
6. Proposed Frameworks: Bridging the Gap
To move beyond the limbo, Nigeria must transition from “Passive Consumer” to “Active Architect” of AI law.16 The following frameworks are proposed:
6.1 Mandatory Algorithmic Disclosure
NITDA, using its powers under the 2026 Bill, should mandate that any AI service operating in Nigeria must provide an “IP Audit Trail” showing the sources of its training data. This would operationalise transparency obligations similar to those under the EU AI Act, giving Nigerian creators the visibility to assert their rights where their works have been harvested without consent.
6.2 The “Arrangements” Logic Extension
Just as Section 108 of the Copyright Act 2022 grants ownership of a film to the person who “made the arrangements” for its making, the law could be amended to grant limited, sui generis rights to the human who “made the arrangements” — that is, engineered the prompts — for an AI-generated work, provided a high threshold of “creative direction” is met. This would not conflate the human prompter with a traditional author, but would create a defined category of protected interest appropriate to the realities of algorithmic creativity.
Conclusion: Towards a Unified Framework for Algorithmic Integrity
The “Legal Limbo” of the AI Commons is not merely a theoretical debate for the classroom; it is a burgeoning crisis for the Nigerian digital economy. As this article has demonstrated, the intersection of Artificial Intelligence and Intellectual Property is currently a landscape of “unregulated harvesting.” While the Nigerian Copyright Act 2022 provides a modern shield for human creators, it remains vulnerable to the sheer scale of machine-driven reproduction. Simultaneously, the National Digital Economy and E-Governance Bill 2026 offers a robust skeletal structure for a digital Nigeria, yet it lacks the “musculature” of intellectual property protections necessary to safeguard the country’s most valuable asset: its data.
The “Algorithmic Author” should not be viewed as an enemy of the law, but as a technological entity that requires a new category of legal governance. To bridge the gap, the National Assembly must move beyond administrative regulation and embrace Legislative Harmony. This requires a deliberate alignment between the National Information Technology Development Agency (NITDA) and the Nigerian Copyright Commission (NCC). If the 2026 Bill is signed into law without a clear mandate for “IP Transparency” and “Database Sovereignty,” Nigeria risks becoming a digital jurisdiction where innovation flourishes at the expense of its own citizens’ creative rights.
Ultimately, the goal of AI regulation should be to foster an environment where technology serves humanity, not where it replaces the human incentive to create. By amending the current Bill to include mandatory disclosure of training data and recognising the “Human Arrangement” behind AI outputs, Nigeria can lead the African continent in creating a balanced, fair, and future-proof digital ecosystem. The time to act is now, while the 10th National Assembly still holds the pen; for once the algorithm is allowed to write the rules, the human author may find themselves permanently erased from the ledger of ownership.
Footnote(S):
1 Goldsmith LLP, ‘AI and IP: Navigating Authorship & Ownership in Generative Content in Nigeria’ (9 December 2025) https://www.goldsmithsllp.com/ai-and-ip-navigating-authorship-ownership-in-generative-content-in-nigeria/ accessed 18 March 2026.
2 Nigerian Copyright Act 2022, s 2(2)(a).
3 National Digital Economy and E-Governance Bill 2025 (Nigeria), Long Title.
4 Nigerian Copyright Act 2022, s 108.
5 Ibid, s 9(1)(a).
6 Ibid, s 20.
7 See Suswam v. Egwu (2016) 1 NWLR (Pt. 1493) 241, regarding the literal rule of statutory interpretation in Nigeria.
8 Nigerian Copyright Act 2022, s 2(2)(a).
9 National Digital Economy and E-Governance Bill 2025/2026 (Nigeria), Explanatory Memorandum.
10 Ibid, Part I, s 1.
11 Ibid, Part IX, s 42.
12 Ibid, Part XII, s 58.
13 Regulation (EU) 2024/1689 (The EU AI Act), Art 53.
14 Ibid, Recital 107.
15 Getty Images (US), Inc. v. Stability AI, Inc., No. 1:23-cv-00135 (D. Del. 2023).
16 Kashifu Inuwa Abdullahi, ‘NITDA DG Reaffirms FG Commitment to Responsible and Inclusive AI’ (InnovateAI Conference, Lagos, February 2026).
