Albakjaji & Almarzouqi, “The Dilemma of the Copyrights of Artificial Intelligence: The Case of Saudi Arabia Regulations”

by Stephan Kinsella on October 3, 2025

Mohamad Albakjaji and Reem Almarzouqi, “The Dilemma of the Copyrights of Artificial Intelligence,” International Journal of Sociotechnology and Knowledge Development (IJSKD) 16, no.1 (2024): 1-15 (official)

Abstract:

Artificial intelligence (AI) and intellectual property (IP) share some key similarities, such as uncertainty in predictions, processing a massive amount of data, and machine learning. Yet, they also differ from each other. This paper provides background information on how these two domains have evolved over time. It also highlights how Saudi Arabia’s IP system differs from those of other countries. Furthermore, this article explores the relationship between AI and IP and their application in copyright. This study is significant as it helps identify the challenges and opportunities that AI presents with respect to IP in terms of copyright. Finally, this article makes recommendations that will help protect both AI and IP.

Grok summary:

Summary of the Article: “The Dilemma of the Copyrights of Artificial Intelligence: The Case of Saudi Arabia Regulations”

This 2024 open-access article, published in the International Journal of Sociotechnology and Knowledge Development (Volume 16, Issue 1), explores the evolving intersection of artificial intelligence (AI) and intellectual property (IP), with a primary focus on copyright challenges. Authored by Mohamad Albakjaji (Prince Sultan University) and Reem Almarzouqi (Tawal Telecommunications Towers Company), it uses a doctrinal research methodology, analyzing Saudi Arabian laws, international frameworks, case studies, and expert interviews to identify gaps in protecting AI-generated works. The study draws on primary sources (e.g., court rulings, interviews) and secondary sources (e.g., WIPO, WTO documents, academic literature) to argue that current IP systems, including Saudi regulations, are ill-equipped for AI’s unique properties, such as autonomous creation without human-like originality or intent.

The article’s core thesis is that AI’s rapid advancement creates a “dilemma” in copyright: AI can generate protectable works (e.g., art, literature, inventions), but traditional laws tie protection to human authorship, leading to uncertainties in ownership, protection, and enforcement. It highlights Saudi Arabia’s role in AI development under Vision 2030, emphasizing the need for policy reforms to balance innovation with IP rights. While the paper touches on patents and other IP forms (e.g., trademarks, trade secrets), its emphasis is on copyright, noting AI’s potential to process massive data and learn autonomously, which blurs lines between tool and creator.

Key Sections and Findings

Introduction and Background: Discusses AI as a digital transformation strategy impacting fields like copyright. AI-generated works (e.g., artwork, novels) may qualify for IP protection if “original,” but AI lacks human traits like creativity or imagination. The paper contrasts AI with traditional IP, where ownership defaults to the human creator (author, inventor, or designer). It identifies global challenges, such as outdated laws failing to address AI’s independence, and positions Saudi Arabia as a case study for adapting IP frameworks.
Literature Review: Reviews copyright history (e.g., U.S. origins in 1790, WIPO definitions) and distinctions between meanings (legal rules, property rights, remuneration). It explains protection scopes (exclusive rights to reproduce/perform works) and variations (e.g., 70 years post-death in many countries vs. 50 years in Saudi Arabia). The review critiques how AI complicates authorship, citing scholars like Trpathi and Ghatak (2018) on patent-AI interactions.
Copyright in the Era of AI: Analyzes historical cases (e.g., 1884 Burrow-Giles vs. Sarony on photography as a tool) to argue AI could be seen as a human extension or independent actor. U.S. law requires human authorship, excluding AI-generated works unless under “work-for-hire” principles. The paper debates whether AI outputs qualify as “original intellectual creations,” noting denials in jurisdictions like the U.S. and EU.
Copyright Law in Saudi Arabia: Details the Saudi Authority for Intellectual Property (SAIP) and Royal Decree No. M/41 (2003), which protects works like books, art, software, and derivatives for 50 years (or 25 for applied arts/photographs). Exclusions include official documents, news, and abstract ideas. Saudi affiliations (e.g., Berne Convention since 2004) align it with global standards, but the law assumes human authorship.
Challenges of AI and Ownership in Saudi Law: AI’s machine learning (e.g., neural networks generating art/music without human input) challenges human-centric rules. Examples include Google’s AI writing news or “The Next Rembrandt” project. Without protection, AI works could enter the public domain, deterring investment (e.g., millions spent on music-generating AI only for outputs to be freely reusable). The paper contrasts approaches: U.S./EU deny non-human copyright; UK/India attribute to the “arranger” (programmer). It warns of commercial repercussions, like reduced incentives for AI developers, and calls for reforms to clarify ownership (e.g., programmer vs. user).

The article concludes with recommendations for Saudi policy updates to protect AI works, fostering innovation under Vision 2030, while addressing gaps like undefined AI authorship.

Relevance to How Copyright and Patents Pose a Threat to AI

The article primarily frames copyright as a double-edged sword: it risks underprotecting AI outputs (threatening investment and innovation) due to human-authorship requirements, potentially stifling Saudi AI growth. Patents are mentioned briefly as another IP form AI could infringe or generate (e.g., AI-invented medicines), but the focus is on copyright’s hurdles, such as originality thresholds and ownership disputes, which could harm the creative economy by creating legal ambiguities.

Integrating insights from consulted sources (from c4sif.org, a libertarian-leaning site critical of IP), the threats extend beyond output protection to AI’s foundational processes like training and data access:

Threat Category	Description from Article	Broader Threats from Sources
Legal Uncertainties in Authorship and Protection	AI works may not qualify as “original” without human input, leading to denials (e.g., U.S. Copyright Office requires human creators). This could reduce AI investment if outputs aren’t protectable.	Copyright creates a “law wall” limiting AI capabilities (e.g., Grok admits it can’t store/use full texts, restricting detailed responses). Patents add a “thicket” of navigation, questioning their net benefit to innovation.
Restrictions on Training Data and Data Scraping	Not directly addressed, but implied in AI’s data-processing needs; laws may not account for AI learning from existing works.	Copyright hinders scraping public data (e.g., lawsuits like NYT vs. OpenAI over article use; Anthropic’s $1.5B+ settlements). This risks trillions in liabilities, handicapping U.S. AI vs. less-restrictive competitors like China.
Fair Use Debates and Infringement Risks	AI training might be “transformative” (e.g., like photography tools), but uncertainties persist, especially for derivatives.	Fair use is inconsistently applied (e.g., U.S. vs. UK’s limited exceptions); treating training as infringement conflates it with “stealing,” leading to absurd outcomes like pre-emptive censorship.
Economic and Innovation Impacts	Without reforms, AI could undermine IP incentives, reducing developer motivation (e.g., unprotected game music).	IP stifles competition; calls to abolish laws (e.g., Musk/Dorsey) argue they ignore realities like AI deriving works independently, potentially rendering protections obsolete and harming progress.
Global/Saudi-Specific Threats	Saudi laws (50-year terms) assume human authors; gaps could hinder Vision 2030 AI goals.	U.S.-centric lawsuits amplify risks; sources advocate shortening terms (e.g., to 5 years) or abolition to liberate AI, echoing article’s reform calls.

Overall, the article views copyright/patents as outdated barriers that threaten AI by creating ownership voids and investment deterrents, while sources amplify this as systemic “gimping” of AI through data restrictions and lawsuits, advocating radical reforms or abolition to unleash innovation. Two sources provided insufficient content for deeper analysis.