AI companies are under a tsunami of lawsuits for the unauthorised use of copyrighted materials to train their AI models. Authors, artists and publishers in the United States of America, France, Australia, Canada, and India have taken legal actions, marking a global reckoning over intellectual property rights in the age of AI. The New York Times has sued Peperlexity, the artificial intelligence start-up that has built a cutting-edge internet search engine, and OpenAI for training their models on NYT content without authorisation.
The largest legal action to date resulted in a $1.5bn settlement, after a coalition of authors sued Anthropic for using nearly 200,000 copyrighted books to train its large language models without consent or compensation. Court filings revealed that a significant portion of the training data used by Anthropic – as well as Meta and Bloomberg – was drawn from pirated book databases such as Books3 and LibGen, exposing just how deeply unauthorised content has been embedded in the development of frontier AI systems.
The authors whose fiction and non-fiction books were used to train these foundational AI models without consent included globally celebrated African writers, such as South African novelist Zakes Mda; Nigerian Nobel laureate Wole Soyinka; the internationally renowned Chinua Achebe; Nigerian author Chimamanda Ngozi Adichie; Moroccan-American writer Laila Lalami; and Ghanaian literary icon Ama Ata Aidoo, among others. Some of these African writers whose publishing rights were managed by houses in the US, UK, and France are in line to be compensated as part of Anthropic’s $1.5bn settlement.
But there is a striking global and regional regulatory silence by African copyright and competition regulators and media when it comes to the unauthorised use of copyrighted materials written by African authors and published by African-based publishing houses, many of which were scraped and used to train frontier AI models without permission.
Although works by young and emerging African writers, including Damilare Kuku and Billy Kahora, appeared in the pirated databases, they did not qualify for the compensation, which is approximately $3,000 per work,, because their books were not registered with the United States Copyright Office – a requirement under US copyright laws for foreign works from countries that have relevant treaties with the US.
This new challenge raises critical questions about the current state of copyright laws in Africa and the US, and how they can protect local and foreign creatives from intellectual property theft and exploitation in the AI era. Currently, copyright laws in Nigeria and South Africa do not make provisions to sue in the local court for copyright infringements that occurred in a foreign country because of territorial limitations.
Africa’s share of copyright or royalty income or collections is less than 1%, the lowest across the globe. The unauthorised distribution and infringement of music and films, books, and software accounts for up to 40% in lost income for Nollywood alone, one reason Nigerian filmmakers receive only a fraction of the total revenue generated by their movies. The weak, outdated, or unenforceable legal protections for creative content produced and distributed on digital platforms have failed to curb the piracy and counterfeiting threats to the creative industry in Nigeria.
While AI-powered digital platforms have enabled performers, artists, musicians, and others to reach wider audiences, expand into different markets, and increase revenue, AI’s double-edged sword exposes them to a greater risk that multi-billion-dollar AI companies will use their work without compensating them for their efforts.
In Africa, the proliferation of social and entertainment digital platforms, such as Facebook, YouTube, Spotify and others, pushed many countries to amend their copyright laws to protect creators' work from being exploited or distributed online without their consent.
Nigeria’s Copyright Act of 2022, Kenya’s Copyright Act of 2001 and Ghana’s Copyright (Amendment) Act, 2009 all strengthen protections for digital content and audiovisual works by introducing tougher penalties for infringement and empowering creators to ask internet service providers to block access to offending sites. Yet none of these updated copyright laws directly address one of the most urgent new challenges facing creators today: the use of unauthorised materials to train AI models or the legal status of AI-generated content, which raises concerns of reproduction and derivative infringement under the Copyright Act.
South Africa is pioneering a solution to address the challenge by amending its existing copyright law to align with international standards, incorporating provisions that prioritise accessibility, fair use, and the protection of creators' rights from tech companies. But Big Tech’s growing influence raises concerns about how a provision in the bill for “fair use” of copyright materials might open the door for AI companies to train their AI generative systems on South African creators’ work without authorisation or compensation, effectively reproducing the same exploitation risks African creators have long faced in global digital markets.
Big Tech in Africa: Beyond reproach?
Big Tech companies with large numbers of users in Africa, such as Google, Meta and OpenAI, often insist they are neither domiciled nor resident in African countries to argue that they are beyond the reach of local regulators. They position themselves as champions of Africa’s digital transformation, while deploying extensive legal resources to sidestep regulatory scrutiny on the continent.
Although the companies use their significant lobbying and legal resources to evade scrutiny in the EU and the US, their actions in Africa are of a different magnitude.
For example, Meta is still embroiled in a legal battle with Kenyan content moderators who allege unfair treatment and low wages. Yet, in the US, the company paid $52m in settlement to American content moderators who suffered psychological harm due to exposure to traumatic materials they had viewed.
The company also threatened to shut down Nigerian access to Facebook and Instagram, the two biggest social platforms that micro-entrepreneurs in Nigeria rely on for business, when the Federal Competition and Consumer Protection Commission imposed a $290m fine for “invasive practices against data subjects” in the country. Meanwhile, the company paid a $1.4bn settlement to settle a privacy lawsuit in Texas.
African copyright frameworks need revisions
With projections indicating a potential 20% revenue loss for creators in the music and audiovisual sectors due to AI-generated outputs, a critical question arises: can locally amended copyright laws be effectively enforced against foreign AI companies, often operating outside national jurisdictions, when they use unauthorised African IP works to reproduce songs or generate images trained on datasets containing local photographers’ material? To address this growing threat, African regulators must strengthen existing copyright frameworks to include explicit provisions that address AI-generated works, including human authorship requirements, ownership, and originality standards to protect African creators from potential infringement by both local and foreign AI companies.
With the rise in technology-related legal disputes, Africa needs to establish standalone regional technology courts dedicated exclusively to handling digital and AI-related cases, similar to the Guangzhou Internet Court in China. At the same time, existing regulatory institutions must build the capacity to detect and respond to AI-driven intellectual property violations, and develop mechanisms to compensate IP holders whose work is used to inspire or generate AI content.
There is also an urgent need to raise public awareness within the creative industries about AI and its implications for intellectual property rights, including the creation of AI-creative explainability guidelines. African creators must understand how AI systems can infringe on their work, the technical skills needed to identify and assess such infringements, and the legal pathways available for seeking remedies.
