The Growing Push for AI Transparency: What South African Businesses Should Know

Artificial intelligence has moved beyond the hype and is now delivering real business value, as companies across the world use it to generate marketing material, social media content, product images, videos and customer communications.

As synthetic content becomes indistinguishable from authentic content, ordinary people lose the ability to make informed judgements about what they’re seeing, hearing, or interacting with, unless the law forces disclosure.  In the European Union (EU) certain AI-generated or AI-manipulated content is required to be disclosed, labelled and capable of being traced back to its source.

The objective is not to prevent the use of AI, but to ensure transparency and maintain consumer trust.  For businesses with partners in Europe, these requirements have immediate implications. As South Africa moves toward its own AI policy framework, local businesses may in time face the same or similar compliance requirements in respect of transparency, accountability and the traceability of AI-generated content.

The Current Legal Position in South Africa

South Africa does not currently have a dedicated AI regulatory framework, nor is there any general legal requirement to label AI-generated advertisements, social media posts, marketing flyers or website content.

There has been considerable discussion around AI governance, including the publication of a Draft National AI Policy earlier in 2026. The policy was subsequently withdrawn after concerns emerged regarding the integrity of certain references included in the document, and no replacement policy has yet been issued.

For now, businesses remain largely free to use AI-generated content without formally disclosing it. That does not mean, however, that AI operates outside existing legal frameworks.

Many of the risks associated with AI-generated content are already addressed by existing legislation and common law. The question is often not whether AI was used, but whether the resulting content infringes rights, misleads consumers or creates a false impression.

The Consumer Protection Act 68 of 2008 prohibits suppliers from making false, misleading or deceptive representations. If a business uses AI-generated images that exaggerate the appearance, quality or functionality of a product, the fact that AI created the image is unlikely to be relevant. The focus will be on whether the consumer was misled.

The same principles apply to AI-generated testimonials, fabricated reviews, manipulated product images and synthetic endorsements.

Similarly, the Advertising Regulatory Board’s Code of Advertising Practice requires advertising to be truthful and not likely to mislead consumers. Although the Code was not drafted with generative AI in mind, its principles apply just as readily to AI-generated content.

Many of the risks associated with AI-generated content do not arise because AI was used. They arise because businesses fail, on the one hand, to consider how the content will be perceived by consumers, competitors and regulators and, on the other hand, fail to undertake due diligence to assess whether an AI tool may have generated content that infringes the rights of a third party.

AI-generated content can inadvertently mislead consumers. AI-generated images used in cosmetic commercials, for example, may create an unrealistic impression of a product’s effects. Music, too, can be generated using the distinctive style, sound, voice and other elements of a human artist’s work.

The Brand and IP Risks

For brand owners, AI-generated content raises additional concerns.

Businesses are increasingly using AI to generate advertising and media content, reducing costs and increasing efficiency. This convenience, however, can create uncertainty regarding ownership, originality and infringement.

Businesses should consider whether AI-generated content:

• incorporates elements that may infringe another party’s intellectual property rights;
• mimics a competitor’s branding or trade dress;
• creates uncertainty regarding copyright ownership;
• undermines brand distinctiveness by relying on generic AI-generated imagery;
• exposes the business to claims relating to passing off or unlawful competition;

• uses personal information in violation of applicable legislation.

As AI-generated content becomes more prevalent, these questions are likely to become increasingly important from both an IP and risk-management perspective.

What Businesses Should Be Doing Now!

Although the withdrawn Draft National AI Policy is no longer in force, its content offers a useful indication of the direction South African policy is likely to take. In the absence of a dedicated domestic AI regulatory framework, businesses should also have regard to international developments, particularly the European Union’s AI regulatory regime, which provides valuable guidance on emerging best practices, transparency requirements and governance standards for the responsible use of AI.

Businesses should, in the interim, ensure that their use of AI accounts diligently for consumer protection, privacy and digital governance.

The more immediate consideration for businesses is not whether a specific AI labelling law exists today. It is whether their use of AI creates legal, intellectual property, reputational or consumer trust risks.

The organisations best positioned for the future will be those that implement appropriate oversight, governance and review processes now, rather than waiting for regulation to catch up.

At Adams & Adams, we are increasingly assisting clients to assess the legal and commercial risks associated with AI-generated content, including trade mark, copyright and brand protection issues, advertising and consumer protection considerations, as well as broader governance and compliance concerns. As the use of AI becomes more widespread, businesses should ensure that innovation is supported by appropriate oversight, review and risk management processes.

For more information, please reach to the authors at melissa.morris@adams.africa or jameel.hamid@adams.africa.