In Ghana and across Africa, brands are confronting immediate threats and opportunities presented by generative artificial intelligence (AI), particularly concerning its integration into branding and advertising. This evolving landscape challenges established norms around intellectual property, liability, and regulation, as AI models are trained on copyrighted content and generate outputs that may infringe on protected marks or depict natural persons without consent. The intersection of AI technology with branding necessitates a critical examination of ownership, infringement risks, and accountability frameworks for AI-generated content.
The Input Stage: Copyright and AI Training Data
Generative AI models learn from vast datasets, often scraped from the internet without explicit permission. This practice raises significant legal questions about whether unauthorized use of copyrighted works for AI training constitutes fair use or copyright infringement.
Recent legal developments highlight this tension. In 2025, a U.S. court ruled in Bartz v. Anthropic that legally sourced materials could be used for AI training under fair use, but pirated books did not qualify, leading to a $1.5 billion settlement requiring Anthropic to destroy infringing datasets. Another U.S. court suggested Meta’s use of copyrighted material might be transformative but stressed the need for compensation to copyright holders.
In the UK, Getty Images v Stability AI saw Getty largely unsuccessful on copyright claims but achieved a narrow ruling on trademark infringement. This decision offers interim protection for AI developers regarding training data, provided they avoid reproducing proprietary watermarks in outputs.
Nigeria’s Advertising Regulatory Council (ARCON) has warned digital marketers about AI-generated advertisements containing deceptive health claims, posing public health risks. Brands should regularly review how their copyrighted materials might be used in AI training datasets and actively protect their intellectual property rights through take-down requests.
The Output Stage: Ownership and Infringement of AI-Generated Content
The question of ownership for AI-generated brand assets is increasingly being clarified. In March 2026, the U.S. Supreme Court declined to review Thaler v. Perlmutter, affirming that purely AI-generated works without significant human creative input are ineligible for copyright protection.
This means logos, ad copy, or product descriptions created solely by AI for a brand will not receive copyright protection. However, U.S. courts are becoming more receptive to output-based infringement claims, where an AI’s output directly violates existing copyrights. This distinction is crucial, as AI creators may be liable if their systems reproduce copyrighted material owned by others, even if they cannot claim copyright themselves.
Ghanaian brands should meticulously document human creative contributions to any AI-assisted work. Relying exclusively on AI for core brand assets carries significant legal risks.
Trademarks in the Age of AI
Trademark law, based on commercial use, faces new challenges with AI. Liability can arise when trademarks are used in training data, when AI-generated content incorporates protected marks, or when an AI product is named in a way that conflicts with an existing trademark.
Cameo v. OpenAI illustrates this risk. Cameo sued OpenAI over its ‘cameo’ feature in its Sora video app, alleging consumer confusion due to the name’s similarity to Cameo’s established platform. The court issued a temporary injunction, highlighting the legal dangers of adopting well-known brand names for AI products.
In the UK’s Getty Images case, Stability AI was found liable for using watermarks similar to Getty’s in AI outputs. The court determined that Stability AI’s control over training, prompts, and results met the ‘use in the course of trade’ standard, and technical autonomy did not absolve the provider of responsibility.
For Ghanaian brands, watermarking digital assets is more than aesthetic; it serves as crucial evidence in infringement actions. Comprehensive trademark clearance for all AI-generated brand names, slogans, and logos before deployment is essential.
Allocating Liability for AI-Generated Content Violations
Liability for AI content that violates intellectual property rights can fall on the AI developer, the brand/advertiser, or the platform distributing the output.
AI Developer Liability
The Getty Images case established that an AI developer can be liable for trademark infringement if they exercise sufficient control over their outputs. Factors include control over training materials, ability to filter prompts and outputs, user inability to control core model behavior, and consumer perception of the platform as the originator.
Brand/Advertiser Liability
Emerging brands and startups face immediate risks if their AI-generated advertisements fail to comply with existing legal frameworks. AI-generated content mimicking real endorsers, celebrities, or testimonials could breach consumer protection laws and advertising codes in Ghana.
Brands should develop a simplified compliance checklist for AI-generated content in advertising to mitigate litigation risks and ensure adherence to legal standards.
Platform Liability and Safe Harbours
The EU’s Artificial Intelligence Act (AI Act) mandates transparency for AI-generated advertisements, including synthetic media, requiring clear and distinguishable disclosure. Non-compliance can lead to substantial fines.
A reported instance of potential non-compliance involved a Guess advertisement in Vogue magazine, which featured AI-generated imagery with a discreet disclaimer. If this ad circulates in Europe, Guess could face enforcement actions, even if the creative work originated outside the EU. Brands expanding into European markets must adhere to these regulations.
The African and Ghanaian Regulatory Response
Ghana currently lacks a dedicated AI governance statute but is positioning itself as a leader in responsible AI governance through policy frameworks.
The Ghana AI Practitioners’ Guide (GAIPG), introduced in September 2025, provides a framework for ethical AI development tailored to Ghana’s context, emphasizing fairness, accountability, and transparency.
Stakeholder consultations are underway for the Emerging Technologies Bill, which aims to establish an Emerging Technologies Agency with a specialized AI Division for systematic oversight of AI systems.
The forthcoming Data Protection Bill is intended to repeal the current Data Protection Act and amend the Electronic Transactions Act to better address AI challenges, including automated decision-making, deepfakes, and cross-border data transfers. This bill responds to concerns about data harvesting by multinational corporations and the lack of domestic data infrastructure.
Ghanaian courts may look to similar foreign decisions for guidance in the absence of local precedent on AI-generated ad infringements. Across the continent, Nigeria’s ARCON is tackling fraudulent AI ads, and other nations like Kenya, South Africa, and Rwanda are advancing their AI initiatives. The African Continental Free Trade Area (AfCFTA) will necessitate harmonized AI regulations.
Practical Recommendations for Brands
To mitigate risks associated with AI-generated advertisements, brands should adopt several key strategies:
- IP Creation: Document human creative input in AI-assisted works. Avoid sole reliance on AI for core brand assets.
- Training Data: Monitor the use of copyrighted materials in third-party AI model training and explore opt-out or licensing options.
- Trademark Use: Conduct thorough trademark clearance for all AI-generated brand elements.
- Advertising Compliance: Clearly and prominently identify AI-generated advertisements to consumers.
- Vendor Contracts: Update contracts to assign liability for IP infringement from AI outputs.
- Regulatory Monitoring: Stay informed about Ghana’s Emerging Technologies Bill and Data Protection Bill.
- Consumer Protection: Avoid using AI content that could mislead consumers about authenticity or performance.
The advent of AI does not negate existing legal and ethical standards. Laws like copyright, trademark, and consumer protection remain relevant, with AI introducing new avenues for infringement. Ghana’s approach, guided by the GAIPG and upcoming legislation, seeks to balance innovation with responsibility. Brands that integrate compliance measures, document human contributions, review AI content for IP violations, and provide clear disclosures can mitigate legal exposure and build consumer trust in an AI-driven digital landscape. As technology advances, fundamental legal principles of fairness, authorship, and accountability remain paramount.
Leave a Reply